Sentiment Analysis using RNN _ LSTM (ULMFiT).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "8f4ba61032210e7c6c38d1e10b2945d3f7ff5f6d"
   },
   "source": [
    "# Twitter Airline Sentiment Analysis (ULMFiT)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "002138cc3d30e60595f6fdd0249b5274f69645d4"
   },
   "source": [
    "## Introduction"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "db8eaed4bba95cbf0d6e51e4af3fab888dec6867"
   },
   "source": [
    "This notebook explores the [Twitter US Airline Sentiment Dataset](https://www.kaggle.com/crowdflower/twitter-airline-sentiment) and tries to predict tweet sentiment using a language model and RNN via Fast.ai's library for Universal Language Model Fine-tuning for Text Classification ([ULMFiT](http://nlp.fast.ai/classification/2018/05/15/introducting-ulmfit.html))."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "ae1c4edfe924fc366ebfc7d7338434f18b948296"
   },
   "source": [
    "## Setup"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "e0893914ffed12ac6308cb6bfbfe8dc177ebada5"
   },
   "source": [
    "Importing packages."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_uuid": "faca2bd9f6756ed19bd12d69d68f997a272f3169"
   },
   "outputs": [],
   "source": [
    "# Basic packages\n",
    "import pandas as pd \n",
    "import numpy as np\n",
    "import re\n",
    "import collections\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "# Modeling, selection, and evaluation\n",
    "from fastai.text import *\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn.metrics import confusion_matrix, accuracy_score\n",
    "from sklearn.utils.multiclass import unique_labels\n",
    "\n",
    "%matplotlib inline "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "5f7a8fe5ae9344468d4882136f2d9ce46006397e"
   },
   "source": [
    "## Data Preparation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "a3eb5a1294773d7aa1a72fd7979c625d53fedce2"
   },
   "source": [
    "First we read in the data and have a look at the columns we can use and explore. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "_uuid": "47254dc4d6b6dc5d49438ee45aa91f4327922f6c"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>tweet_id</th>\n",
       "      <th>airline_sentiment</th>\n",
       "      <th>airline_sentiment_confidence</th>\n",
       "      <th>negativereason</th>\n",
       "      <th>negativereason_confidence</th>\n",
       "      <th>airline</th>\n",
       "      <th>airline_sentiment_gold</th>\n",
       "      <th>name</th>\n",
       "      <th>negativereason_gold</th>\n",
       "      <th>retweet_count</th>\n",
       "      <th>text</th>\n",
       "      <th>tweet_coord</th>\n",
       "      <th>tweet_created</th>\n",
       "      <th>tweet_location</th>\n",
       "      <th>user_timezone</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>8043</th>\n",
       "      <td>568841355226517504</td>\n",
       "      <td>neutral</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Delta</td>\n",
       "      <td>NaN</td>\n",
       "      <td>laurafee</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>@JetBlue well I'm not sure I'm that bold! lol or are you saying you didn't believe me?? :P</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-02-20 10:35:22 -0800</td>\n",
       "      <td>so cal</td>\n",
       "      <td>Pacific Time (US &amp; Canada)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1329</th>\n",
       "      <td>569824565091823616</td>\n",
       "      <td>positive</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>United</td>\n",
       "      <td>NaN</td>\n",
       "      <td>TravlnKB</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>@united EWR agent Barbara was FABULOUS and an example of CUST. SERV. A pleasure talking to you😊 http://t.co/KMQuLY9g5E</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-02-23 03:42:17 -0800</td>\n",
       "      <td>East Coast</td>\n",
       "      <td>Quito</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8166</th>\n",
       "      <td>568642357379399681</td>\n",
       "      <td>positive</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Delta</td>\n",
       "      <td>NaN</td>\n",
       "      <td>adatts33</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>@JetBlue You definitely will!</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-02-19 21:24:37 -0800</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6812</th>\n",
       "      <td>570238950801059840</td>\n",
       "      <td>positive</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>Delta</td>\n",
       "      <td>NaN</td>\n",
       "      <td>lalunkee</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>@JetBlue Annnndddd, I just booked my flight. That was easy.</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-02-24 07:08:55 -0800</td>\n",
       "      <td>All over ...</td>\n",
       "      <td>Eastern Time (US &amp; Canada)</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10587</th>\n",
       "      <td>569159165626716160</td>\n",
       "      <td>negative</td>\n",
       "      <td>1.0</td>\n",
       "      <td>Late Flight</td>\n",
       "      <td>1.0</td>\n",
       "      <td>US Airways</td>\n",
       "      <td>NaN</td>\n",
       "      <td>therealcab</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "      <td>@USAirways now the 3rd issue with the plane is a fuel leak. They will let us know soon what they decide. #getmeoffthisplane #ineedabeer</td>\n",
       "      <td>NaN</td>\n",
       "      <td>2015-02-21 07:38:14 -0800</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 tweet_id airline_sentiment  airline_sentiment_confidence  \\\n",
       "8043   568841355226517504  neutral           1.0                            \n",
       "1329   569824565091823616  positive          1.0                            \n",
       "8166   568642357379399681  positive          1.0                            \n",
       "6812   570238950801059840  positive          1.0                            \n",
       "10587  569159165626716160  negative          1.0                            \n",
       "\n",
       "      negativereason  negativereason_confidence     airline  \\\n",
       "8043   NaN           NaN                         Delta        \n",
       "1329   NaN           NaN                         United       \n",
       "8166   NaN           NaN                         Delta        \n",
       "6812   NaN           NaN                         Delta        \n",
       "10587  Late Flight    1.0                        US Airways   \n",
       "\n",
       "      airline_sentiment_gold        name negativereason_gold  retweet_count  \\\n",
       "8043   NaN                    laurafee    NaN                 0               \n",
       "1329   NaN                    TravlnKB    NaN                 0               \n",
       "8166   NaN                    adatts33    NaN                 0               \n",
       "6812   NaN                    lalunkee    NaN                 0               \n",
       "10587  NaN                    therealcab  NaN                 0               \n",
       "\n",
       "                                                                                                                                          text  \\\n",
       "8043   @JetBlue well I'm not sure I'm that bold! lol or are you saying you didn't believe me?? :P                                                \n",
       "1329   @united EWR agent Barbara was FABULOUS and an example of CUST. SERV. A pleasure talking to you😊 http://t.co/KMQuLY9g5E                    \n",
       "8166   @JetBlue You definitely will!                                                                                                             \n",
       "6812   @JetBlue Annnndddd, I just booked my flight. That was easy.                                                                               \n",
       "10587  @USAirways now the 3rd issue with the plane is a fuel leak. They will let us know soon what they decide. #getmeoffthisplane #ineedabeer   \n",
       "\n",
       "      tweet_coord              tweet_created tweet_location  \\\n",
       "8043   NaN         2015-02-20 10:35:22 -0800  so cal          \n",
       "1329   NaN         2015-02-23 03:42:17 -0800  East Coast      \n",
       "8166   NaN         2015-02-19 21:24:37 -0800  NaN             \n",
       "6812   NaN         2015-02-24 07:08:55 -0800  All over ...    \n",
       "10587  NaN         2015-02-21 07:38:14 -0800  NaN             \n",
       "\n",
       "                    user_timezone  \n",
       "8043   Pacific Time (US & Canada)  \n",
       "1329   Quito                       \n",
       "8166   NaN                         \n",
       "6812   Eastern Time (US & Canada)  \n",
       "10587  NaN                         "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Read file into dataframe\n",
    "pd.set_option('display.max_colwidth', -1)\n",
    "df = pd.read_csv('data/Tweets.csv')\n",
    "df = df.reindex(np.random.permutation(df.index))  \n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "8cede8e8c36c138ec54636d12e3d052be9616f35"
   },
   "source": [
    "Next we look at the distribution of the main dependent variable: airline_sentiment, and its breakdown across airlines and tweet length. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "_uuid": "d6297e30091968d61aad648007f558f016643ac2"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x224c616b518>"
      ]
     },
     "execution_count": 3,
     "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": [
    "df['airline_sentiment'].value_counts().plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "_uuid": "2b1187ba0c78ee9c724d1debef911e8f92d12077"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x224c620ed30>"
      ]
     },
     "execution_count": 4,
     "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": [
    "df['airline'].value_counts().plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "_uuid": "76ee0fbd2697f44de81cc6000939792a0ad290dd"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x224c6414710>"
      ]
     },
     "execution_count": 5,
     "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": [
    "df.groupby(['airline', 'airline_sentiment']).size().unstack().plot(kind='bar', stacked=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "_uuid": "20599ba5c6b31a7e2cb04310e0fb2180c18de84e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x224c64b8160>"
      ]
     },
     "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": [
    "df['tweet_length'] = df['text'].apply(len)\n",
    "df.groupby(['tweet_length', 'airline_sentiment']).size().unstack().plot(kind='line', stacked=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "e2dc85e6cfe1af4c8b94616d2b9dfee34cabd3d0"
   },
   "source": [
    "We see that there isn't a lot of correlation between the number of positive / neutral tweets and the tweet length, but for negative tweets the distribution is heavily skewed towards longer tweets. This is possibly because the angier the tweeter, the more they have to say. Next we'll see if the same observation can be made about the average and median sentiment confidence as well."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "_uuid": "eab6234411542ec97b96a1e726db583081ad5798"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x224c6557080>"
      ]
     },
     "execution_count": 7,
     "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": [
    "df[['tweet_length', 'airline_sentiment', 'airline_sentiment_confidence']].groupby(['tweet_length', 'airline_sentiment']).mean().unstack().plot(kind='line', stacked=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "_uuid": "f40c6280580908146eb1cf3c074dd25641cdccb6"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x224c65c3e10>"
      ]
     },
     "execution_count": 8,
     "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": [
    "df[['tweet_length', 'airline_sentiment', 'airline_sentiment_confidence']].groupby(['tweet_length', 'airline_sentiment']).median().unstack().plot(kind='line', stacked=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "c19f751ac9586379cab76159922bb1e6d117ce12"
   },
   "source": [
    "There doesn't seem to be a discernable relationship between the confidence and the tweet length. Next we randomize and split the data, then write to CSVs. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "_uuid": "5593b67c25038a885338cfacc3bc0db2faa59d5b"
   },
   "outputs": [],
   "source": [
    "test_percentage = 0.1\n",
    "df.sort_index(inplace=True)\n",
    "cutoff = int(test_percentage * df.shape[0])\n",
    "df[['airline_sentiment', 'text']][:cutoff].to_csv('Tweets_filtered_test.csv', index=False, encoding='utf-8')\n",
    "df[['airline_sentiment', 'text']][cutoff:].to_csv('Tweets_filtered_train.csv', index=False, encoding='utf-8')\n",
    "df[['text']][cutoff:].to_csv('Tweets_text_only_train.csv', index=False, encoding='utf-8')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "749aad8c97cda59ca3d5cc6e1109c11fba226f03"
   },
   "source": [
    "### Tokenization"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "c4b5c501b1635e1dbe2e998094adf94432cef679"
   },
   "source": [
    "We read in the data and add new words to our dictionary, as well as create a representation of words using numbers. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "_uuid": "eadc113d34dcec35521146b95b876a895f486d3e"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>text</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>xxbos @united xxmaj hi have a question re future xxmaj flight xxmaj booking xxmaj problems . xxup dub - xxup jac 29 / 9 xxup jac - xxup lax 8 / 10 xxup lax - xxup dub 13 / 10 . i 'm * xxup xxunk xxmaj what is checked bag allowance for xxup jac - xxup lax ?</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @united i xxup just xxup asked xxup my xxup boyfriend xxup to xxup prom xxup over xxup the xxup xxunk xxup on xxup flight 494 xxup he xxup said xxup yes xxrep 4 ! xxup best xxup day xxup ever ! ! ! xxup thank u xxup so xxup much xxrep 6 !</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @americanair xxmaj hi . i have xxup koa - xxup lax - xxup phl - xxup ord booked as a xxunk xxunk xxunk . xxmaj if i called to xxunk it to xxup koa - xxup lax - xxup phx - xxup ord would i have to pay any fees</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @americanair xxmaj took a flight yesterday fron xxup tpa - xxup dfw - xxup aus . xxmaj bag went to xxup iah . xxmaj today bag went to xxup iah - xxup dfw and then back to xxup iah . xxmaj still no bag !</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @united : xxmaj unhappy with xxmaj united 's service ? \\n \\n  xxmaj read xxmaj xxunk xxmaj xxunk 's open letter to xxup ual 's xxup ceo . \\n \\n  xxmaj leave a comment or xxup rt . \\n \\n  http : / / t.co / xxunk</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data = TextClasDataBunch.from_csv('.', 'Tweets_filtered_train.csv')\n",
    "data.show_batch()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "790136e685728d6f28086cfcb7ef3bb93c6e6840"
   },
   "source": [
    "The following changes have been made to the text for ease of modeling:\n",
    "- split on space and punctuation symbols\n",
    "- the \"'s\" are grouped together in one token\n",
    "- the contractions are separated like this: \"did\", \"n't\"\n",
    "- there are several special tokens (all those that begin by xx), to replace unknown tokens (see below)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "_uuid": "59520bc3f7302e8556a2e5df45f36f0d5c250632"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['xxunk',\n",
       " 'xxpad',\n",
       " 'xxbos',\n",
       " 'xxeos',\n",
       " 'xxfld',\n",
       " 'xxmaj',\n",
       " 'xxup',\n",
       " 'xxrep',\n",
       " 'xxwrep',\n",
       " '.']"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data.vocab.itos[:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "8f6a97d7683f622c7ed39aab51b0f5c18e22b263"
   },
   "source": [
    "Looking into the dataset we can now see the current representation post-processing, in both text form and numerical."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "_uuid": "6f613b4a36c92a7886732faac4b0558c637cad0d"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "xxbos @usairways thank you for refunding my bag fee . i look forward to its return today i hope .\n",
      "xxbos @americanair i need xxup help ! i 'm trying to xxmaj cancelled xxmaj flight a flight before it takes off but i ca n't get a hold of xxup anyone because all lines are busy .\n",
      "xxbos @united i can understand that but to not have another plane available when there is a full day till departure is unacceptable . xxmaj phoenix ?\n"
     ]
    }
   ],
   "source": [
    "print(data.train_ds[0][0])\n",
    "print(data.train_ds[1][0])\n",
    "print(data.train_ds[2][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "_uuid": "1b42d12d4e9b0c1d046d23a8881318de4daf6ff5"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[   2   24   92   16   19 3883   25  115  539    9]\n",
      "[  2  26  11 101   6  67  13  11  84 142]\n",
      "[  2  27  11  54 366  37  52  10  41  38]\n"
     ]
    }
   ],
   "source": [
    "print(data.train_ds[0][0].data[:10])\n",
    "print(data.train_ds[1][0].data[:10])\n",
    "print(data.train_ds[2][0].data[:10])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "2f91e80464517fde3de71b048f015585212de95b"
   },
   "source": [
    "## Language Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "5b13610fc4e8f88b213f4299c1f567b06711a4e9"
   },
   "source": [
    "We'll be using a language model provided from the fastai library and applying the pre-calculated weights from wikitext-103. This will provide a word embedding scheme that aligns with the corpus of airline tweets and will hopefully capture sufficient \"meaning\" in each word. The language model here will not be optimal as tweets sometimes do not obey conventional spelling and grammar, especially hastags. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "_uuid": "e59b3e84415ea26a2f2eb13ff9ab8d72d13c02da"
   },
   "outputs": [],
   "source": [
    "bs = 24\n",
    "seed = 333"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "577868658a845cb78ec6d2c182c7771865d58242"
   },
   "source": [
    "First we ingest the data once again but have a 10% holdout only since we'll be using it to calibrate the language model. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "_uuid": "eb04883c5fa59017586d8b0b6bed910d9a519f6c"
   },
   "outputs": [],
   "source": [
    "data_lm = (TextList.from_csv('.', 'Tweets_text_only_train.csv')\n",
    "            .random_split_by_pct(0.1, seed = seed)\n",
    "           #We randomly split and keep 10% for validation\n",
    "            .label_for_lm()           \n",
    "           #We want to do a language model so we label accordingly\n",
    "            .databunch(bs=bs))\n",
    "data_lm.save('data_lm.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "_uuid": "51adb9e8db647d136c8b071ff5280bf04bd74332"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>idx</th>\n",
       "      <th>text</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>xxmaj xxunk ( xxmaj australia ) to xxmaj bogota ( xxmaj xxunk ) tomorrow xxbos @united xxmaj hi . xxmaj my relative 's xxmaj flight xxmaj booking xxmaj problems number is xxup xxunk . xxmaj her name is xxmaj xxunk xxmaj xxunk xxmaj xxunk . xxmaj thank you xxbos @united it was delivered ! xxmaj thank you for making sure it arrived at my xxunk ! xxbos @united - rebooked</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>good to know . xxbos @united okay . xxmaj thanks . xxbos @united xxmaj conference xxunk in 3 hours . xxmaj up all night due to delays and still waiting to talk to someone about lost luggage . xxbos @united xxmaj seat xxup xxunk , xxmaj flight xxup xxunk xxbos @united is it on a flight now ? xxmaj thanks for reply . xxbos @united is xxunk and i am</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>and the delay turns into a xxmaj cancelled xxmaj flight ... xxbos @united xxup ua xxunk delayed again - i get charged $ 600 if xxmaj late xxmaj flight by you - you say \" oh well \" xxbos @united we are trying to go as far away from king'scollegelondon as possible for # charity today . xxmaj would you help us ? # jailbreak # xxup rag xxbos @united</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>1st class . xxmaj xxunk ? xxmaj really ? @united : xxunk xxmaj your xxunk is xxunk xxbos @united xxmaj thanks xxmaj yup i 'm all set . xxmaj it happens . xxup slc ground staff were prompt , helpful and courteous . xxbos xxmaj there were plenty of empty seats in coach \" @united : xxmaj your xxunk is understood . xxmaj crew members traveling for duty are pre</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>when i boarded . xxbos “ @united : xxunk xxmaj sorry to hear about your flight . xxmaj do you need help reflight xxmaj booking xxmaj problems ? ” \\n \\n  👎 xxbos @united now we are trying to get to xxmaj san xxmaj juan from xxmaj chicago o'hare . xxmaj having lots of problems . xxmaj may get a standby flight . xxbos @united xxmaj alright , thank</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# data_lm = load_data(path, 'data_lm.pkl', bs=bs)\n",
    "data_lm.show_batch()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "e45e6b553721fe5b9f8cc9b91c05fccf0cb361e5"
   },
   "source": [
    "Now we define the language model and set the learning rates. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "_uuid": "168da9047db23f4c73e325f40a736bad83868b04"
   },
   "outputs": [],
   "source": [
    "learn = language_model_learner(data_lm, AWD_LSTM, drop_mult=0.3)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "_uuid": "9ec59c8b669394b16b104e22cb4bfb48ac6a5e4f"
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR Finder is complete, type {learner_name}.recorder.plot() to see the graph.\n"
     ]
    }
   ],
   "source": [
    "learn.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "_uuid": "8d13823a8ec6e16826c27f29abc90e2a124b05e4"
   },
   "outputs": [
    {
     "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": [
    "learn.recorder.plot(skip_end=15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "4506084188a7f73e55b945108d76d55b8532a848"
   },
   "source": [
    "Next we fit the model for a few cycles by running 1 epoch and then unfreezing and running more epochs to fine tune.  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "_uuid": "9499223dab7f3b70d30c9cfa456de509d5cb9e4e"
   },
   "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>accuracy</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>4.099324</td>\n",
       "      <td>3.806149</td>\n",
       "      <td>0.266131</td>\n",
       "      <td>01:42</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit_one_cycle(1, 1e-2, moms=(0.8,0.7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "_uuid": "591b84005a3cd666f2a8f5b2b01d5a5af764ca60"
   },
   "outputs": [],
   "source": [
    "learn.save('fit_head')\n",
    "# learn.load('fit_head')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "_uuid": "0997782f35a4d3a9fd918377e370d68d2445b7f4"
   },
   "outputs": [],
   "source": [
    "learn.unfreeze()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "_uuid": "7b07d74c4a9aed170c02f34edbb9f3b42fb7dc1b"
   },
   "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>accuracy</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>3.715624</td>\n",
       "      <td>3.670435</td>\n",
       "      <td>0.285952</td>\n",
       "      <td>02:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>3.580942</td>\n",
       "      <td>3.577619</td>\n",
       "      <td>0.295952</td>\n",
       "      <td>02:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>3.408887</td>\n",
       "      <td>3.524817</td>\n",
       "      <td>0.304048</td>\n",
       "      <td>02:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>3.149829</td>\n",
       "      <td>3.522770</td>\n",
       "      <td>0.304435</td>\n",
       "      <td>02:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>2.782777</td>\n",
       "      <td>3.589881</td>\n",
       "      <td>0.302441</td>\n",
       "      <td>02:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>2.406493</td>\n",
       "      <td>3.702429</td>\n",
       "      <td>0.298810</td>\n",
       "      <td>02:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>2.083292</td>\n",
       "      <td>3.842677</td>\n",
       "      <td>0.294494</td>\n",
       "      <td>02:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>1.841232</td>\n",
       "      <td>3.937765</td>\n",
       "      <td>0.293720</td>\n",
       "      <td>02:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>1.674939</td>\n",
       "      <td>3.991176</td>\n",
       "      <td>0.292649</td>\n",
       "      <td>02:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>1.603649</td>\n",
       "      <td>4.013591</td>\n",
       "      <td>0.293333</td>\n",
       "      <td>02:23</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit_one_cycle(10, 1e-3, moms=(0.8,0.7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "_uuid": "ff336814fe0ec59ecc47b855118e655e21298fa5"
   },
   "outputs": [],
   "source": [
    "learn.save('fine_tuned')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "_uuid": "fa1a27351f989101baa799cfc5601f0e31ec428a"
   },
   "outputs": [],
   "source": [
    "learn.save_encoder('fine_tuned_enc')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "2ceac5b715627d84b53f3dc8ff367a713b279b06"
   },
   "source": [
    "The language model is a bit worse than I would have expected; this is likely due to the fact that tweets dont always follow proper English spelling and grammar, making it difficult for the model to correctly predict. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "b28d1b50858cdede3bf149504b9bd30b8e6875f2"
   },
   "source": [
    "## Classifier"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "055671657bcf76f27717c00f4a5de19b5eddf6fe"
   },
   "source": [
    "Next we'll use the encoder from the language model in our classifier, which has a similar LSTM architecture but will predict the sentiment instead of the next word in a tweet. The model architecture here presents some advantages over traditional bags-of-words "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "_uuid": "50144a534af1fca350650121ed4c1a6d1c22376b"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th>text</th>\n",
       "      <th>target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>xxbos @united xxmaj hi have a question re future xxmaj flight xxmaj booking xxmaj problems . xxup xxunk - xxup jac 29 / 9 xxup jac - xxup lax 8 / 10 xxup lax - xxup xxunk 13 / 10 . i 'm * xxup xxunk xxmaj what is checked bag allowance for xxup jac - xxup lax ?</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @americanair xxmaj hi . i have xxup koa - xxup lax - xxup phl - xxup ord booked as a 1-way xxunk xxunk . xxmaj if i called to xxunk it to xxup koa - xxup lax - xxup phx - xxup ord would i have to pay any fees</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @usairways xxmaj my xxmaj flight xxmaj booking xxmaj problems xxup xxunk just times out when i select it under xxmaj manage xxmaj my xxmaj flight xxmaj booking xxmaj problems for months now . i have emailed but no response . xxmaj help ?</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @united ( 2 / 2 ) xxmaj it xxunk that if ca n't confirm xxunk at time of xxmaj flight xxmaj booking xxmaj problems , i should assume that it may never clear . xxmaj did n't used to be that way</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>xxbos @united xxmaj missed xxup xxunk flight due to ur xxmaj cancelled xxmaj flightlation . xxmaj xxunk supervisor disgusting . xxmaj now day xxmaj late xxmaj flight . xxmaj no apology or upgrade offer to xxup lhr from xxup ord ? ? ?</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "data_clas = (TextList.from_csv('.', 'Tweets_filtered_train.csv', cols = 'text')               \n",
    "             .random_split_by_pct(0.1, seed = seed)\n",
    "             .label_from_df(cols=0)\n",
    "             .databunch(bs=bs))\n",
    "data_clas.save('data_clas.pkl')\n",
    "data_clas.show_batch()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "_uuid": "da00b9c3c11ef567092e866c59635d8c3f5f8094"
   },
   "outputs": [],
   "source": [
    "learn = text_classifier_learner(data_clas, AWD_LSTM, drop_mult=0.5)\n",
    "learn.load_encoder('fine_tuned_enc')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "_uuid": "3588b1b9187c1ac8bf02dc58cac6daa872cccde3"
   },
   "outputs": [
    {
     "data": {
      "text/html": [],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "LR Finder is complete, type {learner_name}.recorder.plot() to see the graph.\n"
     ]
    }
   ],
   "source": [
    "learn.lr_find()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "_uuid": "a85a44722e4976b69aaa497bc1fbf7e3f8cfa499"
   },
   "outputs": [
    {
     "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": [
    "learn.recorder.plot()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "0f85bab506edd029007a95cda6916860e2c4a9c0"
   },
   "source": [
    "We train by gradually unfreezing layers and then running an epoch each time, in accordance with the suggestions in the ULMFiT paper."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "_uuid": "667ee3a18340c14dd08ab87a76fbdf569d6bd223"
   },
   "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>accuracy</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.664338</td>\n",
       "      <td>0.523624</td>\n",
       "      <td>0.791192</td>\n",
       "      <td>02:05</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.fit_one_cycle(1, 2e-2, moms=(0.8,0.7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "_uuid": "a2720ee5b3056c080bd22ffe64709aa8187b7809"
   },
   "outputs": [],
   "source": [
    "learn.save('first')\n",
    "# learn.load('first)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "_uuid": "066ec9fa2a5a4d3d37e44c8648a06cf3e83404b8"
   },
   "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>accuracy</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.645647</td>\n",
       "      <td>0.505296</td>\n",
       "      <td>0.802582</td>\n",
       "      <td>02:29</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.freeze_to(-2)\n",
    "learn.fit_one_cycle(1, slice(1e-2/(2.6**4),1e-2), moms=(0.8,0.7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "_uuid": "6107a12891988eda610a53730036b49f5b0d0f2d"
   },
   "outputs": [],
   "source": [
    "learn.save('second')\n",
    "# learn.load('second')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "_uuid": "c35a14796de0fa8da8dc11e37eb2b86742db1412"
   },
   "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>accuracy</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.545377</td>\n",
       "      <td>0.462309</td>\n",
       "      <td>0.821564</td>\n",
       "      <td>03:49</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.freeze_to(-3)\n",
    "learn.fit_one_cycle(1, slice(5e-3/(2.6**4),5e-3), moms=(0.8,0.7))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "_uuid": "60939b1382efb58e5f333ad63bbd022a7b5f075c"
   },
   "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>accuracy</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>0.467728</td>\n",
       "      <td>0.455995</td>\n",
       "      <td>0.821564</td>\n",
       "      <td>04:31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>0.414779</td>\n",
       "      <td>0.485358</td>\n",
       "      <td>0.809415</td>\n",
       "      <td>03:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>0.365361</td>\n",
       "      <td>0.466317</td>\n",
       "      <td>0.817768</td>\n",
       "      <td>03:56</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "learn.unfreeze()\n",
    "learn.fit_one_cycle(3, slice(1e-3/(2.6**4),1e-3), moms=(0.8,0.7))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "480f552d5de6dc7afdb0a3497d1c837fe28e04f5"
   },
   "source": [
    "### Evaluate Performance on Test Set"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "_uuid": "39bf18fc85a5ded5990a8991b8f73fff40fa7843"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x224c8107358>"
      ]
     },
     "execution_count": 36,
     "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": [
    "test_df = pd.read_csv(\"Tweets_filtered_test.csv\", encoding=\"utf-8\")\n",
    "test_df['airline_sentiment'].value_counts().plot(kind='bar')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "_uuid": "a7b63f6e41cbe2acb966e6d5bcf435b1444968b2"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n",
      "C:\\Users\\avi12\\Anaconda3\\lib\\site-packages\\fastai\\torch_core.py:83: UserWarning: Tensor is int32: upgrading to int64; for better performance use int64 input\n",
      "  warn('Tensor is int32: upgrading to int64; for better performance use int64 input')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Test Accuracy:  0.8101092896174863\n"
     ]
    }
   ],
   "source": [
    "test_df['pred_sentiment'] = test_df['text'].apply(lambda row: str(learn.predict(row)[0]))\n",
    "print(\"Test Accuracy: \", accuracy_score(test_df['airline_sentiment'], test_df['pred_sentiment']))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "_uuid": "6516627de2dfa24b6f606adc5da53d826e2c879e"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>airline_sentiment</th>\n",
       "      <th>text</th>\n",
       "      <th>pred_sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>neutral</td>\n",
       "      <td>@VirginAmerica What @dhepburn said.</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica plus you've added commercials to the experience... tacky.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>neutral</td>\n",
       "      <td>@VirginAmerica I didn't today... Must mean I need to take another trip!</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>negative</td>\n",
       "      <td>@VirginAmerica it's really aggressive to blast obnoxious \"entertainment\" in your guests' faces &amp;amp; they have little recourse</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>negative</td>\n",
       "      <td>@VirginAmerica and it's a really big bad thing about it</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>negative</td>\n",
       "      <td>@VirginAmerica seriously would pay $30 a flight for seats that didn't have this playing.\\nit's really the only bad thing about flying VA</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica yes, nearly every time I fly VX this “ear worm” won’t go away :)</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>neutral</td>\n",
       "      <td>@VirginAmerica Really missed a prime opportunity for Men Without Hats parody, there. https://t.co/mWpG7grEZP</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>positive</td>\n",
       "      <td>@virginamerica Well, I didn't…but NOW I DO! :-D</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica it was amazing, and arrived an hour early. You're too good to me.</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>neutral</td>\n",
       "      <td>@VirginAmerica did you know that suicide is the second leading cause of death among teens 10-24</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica I &amp;lt;3 pretty graphics. so much better than minimal iconography. :D</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica This is such a great deal! Already thinking about my 2nd trip to @Australia &amp;amp; I haven't even gone on my 1st trip yet! ;p</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica @virginmedia I'm flying your #fabulous #Seductive skies again! U take all the #stress away from travel http://t.co/ahlXHhKiyn</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica Thanks!</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>negative</td>\n",
       "      <td>@VirginAmerica SFO-PDX schedule is still MIA.</td>\n",
       "      <td>neutral</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica So excited for my first cross country flight LAX to MCO I've heard nothing but great things about Virgin America. #29DaysToGo</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>negative</td>\n",
       "      <td>@VirginAmerica  I flew from NYC to SFO last week and couldn't fully sit in my seat due to two large gentleman on either side of me. HELP!</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>positive</td>\n",
       "      <td>I ❤️ flying @VirginAmerica. ☺️👍</td>\n",
       "      <td>positive</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica you know what would be amazingly awesome? BOS-FLL PLEASE!!!!!!! I want to fly with only you.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   airline_sentiment  \\\n",
       "0   neutral            \n",
       "1   positive           \n",
       "2   neutral            \n",
       "3   negative           \n",
       "4   negative           \n",
       "5   negative           \n",
       "6   positive           \n",
       "7   neutral            \n",
       "8   positive           \n",
       "9   positive           \n",
       "10  neutral            \n",
       "11  positive           \n",
       "12  positive           \n",
       "13  positive           \n",
       "14  positive           \n",
       "15  negative           \n",
       "16  positive           \n",
       "17  negative           \n",
       "18  positive           \n",
       "19  positive           \n",
       "\n",
       "                                                                                                                                            text  \\\n",
       "0   @VirginAmerica What @dhepburn said.                                                                                                            \n",
       "1   @VirginAmerica plus you've added commercials to the experience... tacky.                                                                       \n",
       "2   @VirginAmerica I didn't today... Must mean I need to take another trip!                                                                        \n",
       "3   @VirginAmerica it's really aggressive to blast obnoxious \"entertainment\" in your guests' faces &amp; they have little recourse                 \n",
       "4   @VirginAmerica and it's a really big bad thing about it                                                                                        \n",
       "5   @VirginAmerica seriously would pay $30 a flight for seats that didn't have this playing.\\nit's really the only bad thing about flying VA       \n",
       "6   @VirginAmerica yes, nearly every time I fly VX this “ear worm” won’t go away :)                                                                \n",
       "7   @VirginAmerica Really missed a prime opportunity for Men Without Hats parody, there. https://t.co/mWpG7grEZP                                   \n",
       "8   @virginamerica Well, I didn't…but NOW I DO! :-D                                                                                                \n",
       "9   @VirginAmerica it was amazing, and arrived an hour early. You're too good to me.                                                               \n",
       "10  @VirginAmerica did you know that suicide is the second leading cause of death among teens 10-24                                                \n",
       "11  @VirginAmerica I &lt;3 pretty graphics. so much better than minimal iconography. :D                                                            \n",
       "12  @VirginAmerica This is such a great deal! Already thinking about my 2nd trip to @Australia &amp; I haven't even gone on my 1st trip yet! ;p    \n",
       "13  @VirginAmerica @virginmedia I'm flying your #fabulous #Seductive skies again! U take all the #stress away from travel http://t.co/ahlXHhKiyn   \n",
       "14  @VirginAmerica Thanks!                                                                                                                         \n",
       "15  @VirginAmerica SFO-PDX schedule is still MIA.                                                                                                  \n",
       "16  @VirginAmerica So excited for my first cross country flight LAX to MCO I've heard nothing but great things about Virgin America. #29DaysToGo   \n",
       "17  @VirginAmerica  I flew from NYC to SFO last week and couldn't fully sit in my seat due to two large gentleman on either side of me. HELP!      \n",
       "18  I ❤️ flying @VirginAmerica. ☺️👍                                                                                                                \n",
       "19  @VirginAmerica you know what would be amazingly awesome? BOS-FLL PLEASE!!!!!!! I want to fly with only you.                                    \n",
       "\n",
       "   pred_sentiment  \n",
       "0   neutral        \n",
       "1   negative       \n",
       "2   negative       \n",
       "3   negative       \n",
       "4   negative       \n",
       "5   negative       \n",
       "6   negative       \n",
       "7   negative       \n",
       "8   negative       \n",
       "9   positive       \n",
       "10  neutral        \n",
       "11  positive       \n",
       "12  positive       \n",
       "13  positive       \n",
       "14  positive       \n",
       "15  neutral        \n",
       "16  positive       \n",
       "17  negative       \n",
       "18  positive       \n",
       "19  negative       "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df[:20]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "57beee8541dda946b417f4d9892d2301969ae58c"
   },
   "source": [
    "Plot confusion matrix to see where the areas of misclassification are. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "_uuid": "b9f8e30a2bf5d237cfe6c1f9496ab913bebb5de2"
   },
   "outputs": [],
   "source": [
    "# Confusion matrix plotting adapted from https://scikit-learn.org/stable/auto_examples/model_selection/plot_confusion_matrix.html#sphx-glr-auto-examples-model-selection-plot-confusion-matrix-py\n",
    "def plot_confusion_matrix(y_true, y_pred, classes,\n",
    "                          normalize=False,\n",
    "                          title=None,\n",
    "                          cmap=plt.cm.Blues):\n",
    "    \"\"\"\n",
    "    This function prints and plots the confusion matrix.\n",
    "    Normalization can be applied by setting `normalize=True`.\n",
    "    \"\"\"\n",
    "    if not title:\n",
    "        if normalize:\n",
    "            title = 'Normalized confusion matrix'\n",
    "        else:\n",
    "            title = 'Confusion matrix, without normalization'\n",
    "\n",
    "    # Compute confusion matrix\n",
    "    cm = confusion_matrix(y_true, y_pred)\n",
    "    # Only use the labels that appear in the data\n",
    "    #classes = classes[unique_labels(y_true, y_pred)]\n",
    "\n",
    "    fig, ax = plt.subplots()\n",
    "    im = ax.imshow(cm, interpolation='nearest', cmap=cmap)\n",
    "    ax.figure.colorbar(im, ax=ax)\n",
    "    # We want to show all ticks...\n",
    "    ax.set(xticks=np.arange(cm.shape[1]),\n",
    "           yticks=np.arange(cm.shape[0]),\n",
    "           # ... and label them with the respective list entries\n",
    "           xticklabels=classes, yticklabels=classes,\n",
    "           title=title,\n",
    "           ylabel='True label',\n",
    "           xlabel='Predicted label')\n",
    "\n",
    "    # Rotate the tick labels and set their alignment.\n",
    "    plt.setp(ax.get_xticklabels(), rotation=45, ha=\"right\",\n",
    "             rotation_mode=\"anchor\")\n",
    "\n",
    "    # Loop over data dimensions and create text annotations.\n",
    "    fmt = '.2f' if normalize else 'd'\n",
    "    thresh = cm.max() / 2.\n",
    "    for i in range(cm.shape[0]):\n",
    "        for j in range(cm.shape[1]):\n",
    "            ax.text(j, i, format(cm[i, j], fmt),\n",
    "                    ha=\"center\", va=\"center\",\n",
    "                    color=\"white\" if cm[i, j] > thresh else \"black\")\n",
    "    fig.tight_layout()\n",
    "    return ax"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "_uuid": "2d9d8f505912f65f5e6dc9fc72494f8657204988"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_confusion_matrix(test_df['airline_sentiment'], test_df['pred_sentiment'], classes=['negative', 'neutral', 'positive'], title='Airline sentiment confusion matrix')\n",
    "# confusion_matrix(test_df['airline_sentiment'], test_df['pred_sentiment'], labels=['positive', 'neutral', 'negative'])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "d2494f9841d3bf6ddec89e0ce13478c79d5f7554"
   },
   "source": [
    "From the confusion matrix, we can see that within True Negatives, the prediction accuracy is pretty high. The model has a very hard time classifying neutral tweets, and often misclassifies them as negative. For True Positives, the overall accuracy is pretty good but a surprisingly large number get classified as negative. Let's look at these in closer detail:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "_uuid": "d3d5c67cefb1ea4f932a43dca44665ef65e7243f"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>airline_sentiment</th>\n",
       "      <th>text</th>\n",
       "      <th>pred_sentiment</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica plus you've added commercials to the experience... tacky.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica yes, nearly every time I fly VX this “ear worm” won’t go away :)</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>positive</td>\n",
       "      <td>@virginamerica Well, I didn't…but NOW I DO! :-D</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica you know what would be amazingly awesome? BOS-FLL PLEASE!!!!!!! I want to fly with only you.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica @ladygaga @carrieunderwood  love all three but you really can't beat the classics!</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>81</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica I've applied more then once to be a member of the #inflight crew team...Im 100% interested. #flightattendant #dreampath -G</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>109</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica has getaway deals through May, from $59 one-way. Lots of cool cities http://t.co/tZZJhuIbCH #CheapFlights #FareCompare</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica has getaway deals through May, from $59 one-way. Lots of cool cities http://t.co/B2Xi4YG5T8 #CheapFlights #FareCompare</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>114</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica come back to #PHL already. We need you to take us out of this horrible cold. #pleasecomeback http://t.co/gLXFwP6nQH</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>118</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica your beautiful front-end design is down right now; but it was cool to still book my ticket b/c all your back-end was secure.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>244</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica classiq, luv Virgin America. Greetingz</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>273</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica cutest salt and pepper shaker ever. Just when I think you guys can't get any better you just do! http://t.co/vC6Keulg2J</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>323</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica How about some free drinks on the flight back for the free promo?</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>358</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica just promoting the product is all, had a problem with southwest and recommend noneother than the best! http://t.co/tFaNXBh1Cf</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>369</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica got it squared away. Someone picked up as soon as I tweeted. Should have tweeted sooner. 😉</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>375</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica  Flight Booking Problems last second flight for next week from SFO- to SAN any chance you want to gift me a promo code since I love you guys</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>385</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica that doesn't look to fat to me! It looks yummy!</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>392</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica Your planes are really pretty. Just thought u should know that. :)</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>404</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica this is too cool!  Never been on ur planes but now I will!!</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>421</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica I luv ur people and product! 1 thing is y do u charge for herbal tea but not black tea? I'm not a caffeine drinker</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>455</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica now it's just t-minus 32 minutes until my Elevate a Silver upgrade window opens . #FreeNeverSucks 😃👍</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>459</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica @JezzieGoldz would have been a rough trip but LUCKILY we were on a #virginamerica flight. #weather</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>463</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica another perfect flight.   How come on your planes,  the sun visors can stay down?  Other carriers make you raise them?</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>464</th>\n",
       "      <td>positive</td>\n",
       "      <td>Nice RT @VirginAmerica: The man of steel might be faster, but we have WiFi – just saying. #ScienceBehindTheExperience http://t.co/FGRbpAZSiX</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>480</th>\n",
       "      <td>positive</td>\n",
       "      <td>.@VirginAmerica not only was it great, but you return my calls the day after. Couldn't ask for more. #myVXexperience</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>488</th>\n",
       "      <td>positive</td>\n",
       "      <td>@VirginAmerica so loyal that I'm driving to #NYC from #PA, to fly Virgin,  since you cut #Philly flights ;)</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>527</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united just wanted to let you know how wonderful Rosetta the gate agent was working flight 6457 Dan to Ase. Let her know she wasappreciated</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>552</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united - thanks for your help...got me what I need, but its an issue with @_austrian . I guess I sit awhile longer.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>562</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united Thank you for that. Am I able to claim any interim expenses or is the cost of the stuff up on me?</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>589</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united 441, which also had 1 working WC in coach. Good thing this bird landed ahead of schedule. I have to use the WC stat.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>607</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united despite my bag not making it to Newark good informative email tracking updates help!</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>611</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united Wow. What a deal. Again, 30+ plus seats available. Easy change to make a customer happy.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>663</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united private jet would have been cool! Do does not work. I'll try again</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>690</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united kudos for not Cancelled Flightling flights from DFW this morning.   United usually first to panic...</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>699</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united Every United flight between Saipan and Guam is an adventure! You never know when @CapeAir's old plane will be operational :)</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>721</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united Definitely a compliment! I really thought my bag was lost after it was sent on to another airport. In the end I am a happy customer</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>722</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united thanx so much. You followed through and emailed me a $1000 ticket voucher. #unitedairlines they do care</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>758</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united we just flew with United from AUS (Austin Bergstrom) to Cancun Intl. Best flight ever! The 737's are not too big... Not too small!</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>773</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united flight ua3576, gate b1. And tell ray I somehow made it on to the 736 flight out of IAH and didn't have to wait for the 917 one. :)</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>806</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united you guys continue to impress me in Houston. http://t.co/cIh1qNllcM</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>833</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united Made the upgrade list. Will fly 1st tomorrow (for 40 min) for the first time ever! 🙌 #StatusMatchPaidOff http://t.co/ATfRKp6goY</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>834</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united not just refunded, but for those of us who are on vacation to get a free room night to make up for making us sleep in DIA</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>886</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united captain \"on behalf of the front office, welcome on board.\".... Never heard that before. I laughed...hard.  #funnycaptain</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>941</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united that's exactly the point. It fits. I'm premier access. Boarding group 2. This was a return ticket. I've been doing this for 15 yrs</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>998</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united he has no priority and Iove it</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1036</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united flight 1491...plane from SFO to DEN is basically on time.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1037</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united thanks. Just a program comment; the system assigned seats behind each other even though pairs were available. I changed it back.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1051</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united that would help! or how about integrate it into the App so I can just \"activate\" it and surf...</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1108</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united @JMS2802 : outsource it all United Airlines...your customers  are just loving you for it. Outsource. Outsource. Outsource.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1164</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united would love help getting there today. In #EWR now. Will take any airline and connections. Thx</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1177</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united just confirmed a seat! Crisis averted! Beers won't be missed now</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1315</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united have Michelle at T1 ORD train your other staff on how to treat customers. A refreshing pleasure to deal with.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1352</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united the staff was rather efficient and got us solutions just freaked me out a little being in limbo in the air.</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1415</th>\n",
       "      <td>positive</td>\n",
       "      <td>@united Worked like a charm. Bag was waiting on the carousel when we got to baggage claim. #welldone #goodflight #friendlysky</td>\n",
       "      <td>negative</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     airline_sentiment  \\\n",
       "1     positive           \n",
       "6     positive           \n",
       "8     positive           \n",
       "19    positive           \n",
       "64    positive           \n",
       "81    positive           \n",
       "109   positive           \n",
       "111   positive           \n",
       "114   positive           \n",
       "118   positive           \n",
       "244   positive           \n",
       "273   positive           \n",
       "323   positive           \n",
       "358   positive           \n",
       "369   positive           \n",
       "375   positive           \n",
       "385   positive           \n",
       "392   positive           \n",
       "404   positive           \n",
       "421   positive           \n",
       "455   positive           \n",
       "459   positive           \n",
       "463   positive           \n",
       "464   positive           \n",
       "480   positive           \n",
       "488   positive           \n",
       "527   positive           \n",
       "552   positive           \n",
       "562   positive           \n",
       "589   positive           \n",
       "607   positive           \n",
       "611   positive           \n",
       "663   positive           \n",
       "690   positive           \n",
       "699   positive           \n",
       "721   positive           \n",
       "722   positive           \n",
       "758   positive           \n",
       "773   positive           \n",
       "806   positive           \n",
       "833   positive           \n",
       "834   positive           \n",
       "886   positive           \n",
       "941   positive           \n",
       "998   positive           \n",
       "1036  positive           \n",
       "1037  positive           \n",
       "1051  positive           \n",
       "1108  positive           \n",
       "1164  positive           \n",
       "1177  positive           \n",
       "1315  positive           \n",
       "1352  positive           \n",
       "1415  positive           \n",
       "\n",
       "                                                                                                                                                             text  \\\n",
       "1     @VirginAmerica plus you've added commercials to the experience... tacky.                                                                                      \n",
       "6     @VirginAmerica yes, nearly every time I fly VX this “ear worm” won’t go away :)                                                                               \n",
       "8     @virginamerica Well, I didn't…but NOW I DO! :-D                                                                                                               \n",
       "19    @VirginAmerica you know what would be amazingly awesome? BOS-FLL PLEASE!!!!!!! I want to fly with only you.                                                   \n",
       "64    @VirginAmerica @ladygaga @carrieunderwood  love all three but you really can't beat the classics!                                                             \n",
       "81    @VirginAmerica I've applied more then once to be a member of the #inflight crew team...Im 100% interested. #flightattendant #dreampath -G                     \n",
       "109   @VirginAmerica has getaway deals through May, from $59 one-way. Lots of cool cities http://t.co/tZZJhuIbCH #CheapFlights #FareCompare                         \n",
       "111   @VirginAmerica has getaway deals through May, from $59 one-way. Lots of cool cities http://t.co/B2Xi4YG5T8 #CheapFlights #FareCompare                         \n",
       "114   @VirginAmerica come back to #PHL already. We need you to take us out of this horrible cold. #pleasecomeback http://t.co/gLXFwP6nQH                            \n",
       "118   @VirginAmerica your beautiful front-end design is down right now; but it was cool to still book my ticket b/c all your back-end was secure.                   \n",
       "244   @VirginAmerica classiq, luv Virgin America. Greetingz                                                                                                         \n",
       "273   @VirginAmerica cutest salt and pepper shaker ever. Just when I think you guys can't get any better you just do! http://t.co/vC6Keulg2J                        \n",
       "323   @VirginAmerica How about some free drinks on the flight back for the free promo?                                                                              \n",
       "358   @VirginAmerica just promoting the product is all, had a problem with southwest and recommend noneother than the best! http://t.co/tFaNXBh1Cf                  \n",
       "369   @VirginAmerica got it squared away. Someone picked up as soon as I tweeted. Should have tweeted sooner. 😉                                                     \n",
       "375   @VirginAmerica  Flight Booking Problems last second flight for next week from SFO- to SAN any chance you want to gift me a promo code since I love you guys   \n",
       "385   @VirginAmerica that doesn't look to fat to me! It looks yummy!                                                                                                \n",
       "392   @VirginAmerica Your planes are really pretty. Just thought u should know that. :)                                                                             \n",
       "404   @VirginAmerica this is too cool!  Never been on ur planes but now I will!!                                                                                    \n",
       "421   @VirginAmerica I luv ur people and product! 1 thing is y do u charge for herbal tea but not black tea? I'm not a caffeine drinker                             \n",
       "455   @VirginAmerica now it's just t-minus 32 minutes until my Elevate a Silver upgrade window opens . #FreeNeverSucks 😃👍                                           \n",
       "459   @VirginAmerica @JezzieGoldz would have been a rough trip but LUCKILY we were on a #virginamerica flight. #weather                                             \n",
       "463   @VirginAmerica another perfect flight.   How come on your planes,  the sun visors can stay down?  Other carriers make you raise them?                         \n",
       "464   Nice RT @VirginAmerica: The man of steel might be faster, but we have WiFi – just saying. #ScienceBehindTheExperience http://t.co/FGRbpAZSiX                  \n",
       "480   .@VirginAmerica not only was it great, but you return my calls the day after. Couldn't ask for more. #myVXexperience                                          \n",
       "488   @VirginAmerica so loyal that I'm driving to #NYC from #PA, to fly Virgin,  since you cut #Philly flights ;)                                                   \n",
       "527   @united just wanted to let you know how wonderful Rosetta the gate agent was working flight 6457 Dan to Ase. Let her know she wasappreciated                  \n",
       "552   @united - thanks for your help...got me what I need, but its an issue with @_austrian . I guess I sit awhile longer.                                          \n",
       "562   @united Thank you for that. Am I able to claim any interim expenses or is the cost of the stuff up on me?                                                     \n",
       "589   @united 441, which also had 1 working WC in coach. Good thing this bird landed ahead of schedule. I have to use the WC stat.                                  \n",
       "607   @united despite my bag not making it to Newark good informative email tracking updates help!                                                                  \n",
       "611   @united Wow. What a deal. Again, 30+ plus seats available. Easy change to make a customer happy.                                                              \n",
       "663   @united private jet would have been cool! Do does not work. I'll try again                                                                                    \n",
       "690   @united kudos for not Cancelled Flightling flights from DFW this morning.   United usually first to panic...                                                  \n",
       "699   @united Every United flight between Saipan and Guam is an adventure! You never know when @CapeAir's old plane will be operational :)                          \n",
       "721   @united Definitely a compliment! I really thought my bag was lost after it was sent on to another airport. In the end I am a happy customer                   \n",
       "722   @united thanx so much. You followed through and emailed me a $1000 ticket voucher. #unitedairlines they do care                                               \n",
       "758   @united we just flew with United from AUS (Austin Bergstrom) to Cancun Intl. Best flight ever! The 737's are not too big... Not too small!                    \n",
       "773   @united flight ua3576, gate b1. And tell ray I somehow made it on to the 736 flight out of IAH and didn't have to wait for the 917 one. :)                    \n",
       "806   @united you guys continue to impress me in Houston. http://t.co/cIh1qNllcM                                                                                    \n",
       "833   @united Made the upgrade list. Will fly 1st tomorrow (for 40 min) for the first time ever! 🙌 #StatusMatchPaidOff http://t.co/ATfRKp6goY                       \n",
       "834   @united not just refunded, but for those of us who are on vacation to get a free room night to make up for making us sleep in DIA                             \n",
       "886   @united captain \"on behalf of the front office, welcome on board.\".... Never heard that before. I laughed...hard.  #funnycaptain                              \n",
       "941   @united that's exactly the point. It fits. I'm premier access. Boarding group 2. This was a return ticket. I've been doing this for 15 yrs                    \n",
       "998   @united he has no priority and Iove it                                                                                                                        \n",
       "1036  @united flight 1491...plane from SFO to DEN is basically on time.                                                                                             \n",
       "1037  @united thanks. Just a program comment; the system assigned seats behind each other even though pairs were available. I changed it back.                      \n",
       "1051  @united that would help! or how about integrate it into the App so I can just \"activate\" it and surf...                                                       \n",
       "1108  @united @JMS2802 : outsource it all United Airlines...your customers  are just loving you for it. Outsource. Outsource. Outsource.                            \n",
       "1164  @united would love help getting there today. In #EWR now. Will take any airline and connections. Thx                                                          \n",
       "1177  @united just confirmed a seat! Crisis averted! Beers won't be missed now                                                                                      \n",
       "1315  @united have Michelle at T1 ORD train your other staff on how to treat customers. A refreshing pleasure to deal with.                                         \n",
       "1352  @united the staff was rather efficient and got us solutions just freaked me out a little being in limbo in the air.                                           \n",
       "1415  @united Worked like a charm. Bag was waiting on the carousel when we got to baggage claim. #welldone #goodflight #friendlysky                                 \n",
       "\n",
       "     pred_sentiment  \n",
       "1     negative       \n",
       "6     negative       \n",
       "8     negative       \n",
       "19    negative       \n",
       "64    negative       \n",
       "81    negative       \n",
       "109   negative       \n",
       "111   negative       \n",
       "114   negative       \n",
       "118   negative       \n",
       "244   negative       \n",
       "273   negative       \n",
       "323   negative       \n",
       "358   negative       \n",
       "369   negative       \n",
       "375   negative       \n",
       "385   negative       \n",
       "392   negative       \n",
       "404   negative       \n",
       "421   negative       \n",
       "455   negative       \n",
       "459   negative       \n",
       "463   negative       \n",
       "464   negative       \n",
       "480   negative       \n",
       "488   negative       \n",
       "527   negative       \n",
       "552   negative       \n",
       "562   negative       \n",
       "589   negative       \n",
       "607   negative       \n",
       "611   negative       \n",
       "663   negative       \n",
       "690   negative       \n",
       "699   negative       \n",
       "721   negative       \n",
       "722   negative       \n",
       "758   negative       \n",
       "773   negative       \n",
       "806   negative       \n",
       "833   negative       \n",
       "834   negative       \n",
       "886   negative       \n",
       "941   negative       \n",
       "998   negative       \n",
       "1036  negative       \n",
       "1037  negative       \n",
       "1051  negative       \n",
       "1108  negative       \n",
       "1164  negative       \n",
       "1177  negative       \n",
       "1315  negative       \n",
       "1352  negative       \n",
       "1415  negative       "
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_df.loc[(test_df['airline_sentiment'] == 'positive') & (test_df['pred_sentiment'] == 'negative')]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "c62aa7f9e39908a1455ac0af361d664d54c0e8d8"
   },
   "source": [
    "It's not immediately clear what the common thread that led these tweets to be mislabelled but possible causes include missing signals from hastags, non-grammatical sentences, mis-spellings, etc. which generally contribute to the classification error. \n",
    "\n",
    "In the future, it would be good to spend more time on feature engineering and fitting the language model so that it's more suited to the vocabulary and style of language that is in tweets. From the language model and processing steps, we can see that the model accuracy is not very high and that, even as humans, it's hard to discern the sentiment from the post-processed / tokenized text. One possible way of addressing this issue is to use a much larger corpus of tweets when developing the language model. "
   ]
  }
 ],
 "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": "3.7.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}