{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "ece9e8d9",
   "metadata": {
    "papermill": {
     "duration": 0.102825,
     "end_time": "2022-05-16T23:18:18.905047",
     "exception": false,
     "start_time": "2022-05-16T23:18:18.802222",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Introduction"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0407b5a9",
   "metadata": {
    "papermill": {
     "duration": 0.098538,
     "end_time": "2022-05-16T23:18:19.107274",
     "exception": false,
     "start_time": "2022-05-16T23:18:19.008736",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "One area where deep learning has dramatically improved in the last couple of years is natural language processing (NLP). Computers can now generate text, translate automatically from one language to another, analyze comments, label words in sentences, and much more.\n",
    "\n",
    "Perhaps the most widely practically useful application of NLP is *classification* -- that is, classifying a document automatically into some category. This can be used, for instance, for:\n",
    "\n",
    "- Sentiment analysis (e.g are people saying *positive* or *negative* things about your product)\n",
    "- Author identification (what author most likely wrote some document)\n",
    "- Legal discovery (which documents are in scope for a trial)\n",
    "- Organizing documents by topic\n",
    "- Triaging inbound emails\n",
    "- ...and much more!\n",
    "\n",
    "Classification models can also be used to solve problems that are not, at first, obviously appropriate. For instance, consider the Kaggle [U.S. Patent Phrase to Phrase Matching](https://www.kaggle.com/competitions/us-patent-phrase-to-phrase-matching/) competition. In this, we are tasked with comparing two words or short phrases, and scoring them based on whether they're similar or not, based on which patent class they were used in. With a score of `1` it is considered that the two inputs have identical meaning, and `0` means they have totally different meaning. For instance, *abatement* and *eliminating process* have a score of `0.5`, meaning they're somewhat similar, but not identical.\n",
    "\n",
    "It turns out that this can be represented as a classification problem. How? By representing the question like this:\n",
    "\n",
    "> For the following text...: \"TEXT1: abatement; TEXT2: eliminating process\" ...chose a category of meaning similarity: \"Different; Similar; Identical\".\n",
    "\n",
    "In this notebook we'll see how to solve the Patent Phrase Matching problem by treating it as a classification task, by representing it in a very similar way to that shown above."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72aef3e5",
   "metadata": {
    "papermill": {
     "duration": 0.096761,
     "end_time": "2022-05-16T23:18:19.299885",
     "exception": false,
     "start_time": "2022-05-16T23:18:19.203124",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### On Kaggle"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d7ba1b00",
   "metadata": {
    "papermill": {
     "duration": 0.095382,
     "end_time": "2022-05-16T23:18:19.492836",
     "exception": false,
     "start_time": "2022-05-16T23:18:19.397454",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Kaggle is an awesome resource for aspiring data scientists or anyone looking to improve their machine learning skills. There is nothing like being able to get hands-on practice and receiving real-time feedback to help you improve your skills. It provides:\n",
    "\n",
    "1. Interesting data sets\n",
    "1. Feedback on how you're doing\n",
    "1. A leader board to see what's good, what's possible, and what's state-of-art\n",
    "1. Notebooks and blog posts by winning contestants share useful tips and techniques.\n",
    "\n",
    "The dataset we will be using here is only available from Kaggle. Therefore, you will need to register on the site, then go to the [page for the competition](https://www.kaggle.com/c/us-patent-phrase-to-phrase-matching). On that page click \"Rules,\" then \"I Understand and Accept.\" (Although the competition has finished, and you will not be entering it, you still have to agree to the rules to be allowed to download the data.)\n",
    "\n",
    "There are two ways to then use this data:\n",
    "\n",
    "- Easiest: run this notebook directly on Kaggle, or\n",
    "- Most flexible: download the data locally and run it on your PC or GPU server\n",
    "\n",
    "If you are running this on Kaggle.com, you can skip the next section. Just make sure that on Kaggle you've selected to use a GPU during your session, by clicking on the hamburger menu (3 dots in the top right) and clicking \"Accelerator\" -- it should look like this:"
   ]
  },
  {
   "attachments": {
    "9af4e875-1f2a-468c-b233-8c91531e4c40.png": {
     "image/png": "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"
    },
    "image.png": {
     "image/png": "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"
    }
   },
   "cell_type": "markdown",
   "id": "f7de75fb",
   "metadata": {
    "papermill": {
     "duration": 0.095452,
     "end_time": "2022-05-16T23:18:19.683195",
     "exception": false,
     "start_time": "2022-05-16T23:18:19.587743",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "![image.png](attachment:9af4e875-1f2a-468c-b233-8c91531e4c40.png)!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bfd7f80a",
   "metadata": {
    "papermill": {
     "duration": 0.094882,
     "end_time": "2022-05-16T23:18:19.873521",
     "exception": false,
     "start_time": "2022-05-16T23:18:19.778639",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "We'll need slightly different code depending on whether we're running on Kaggle or not, so we'll use this variable to track where we are:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "a2fd6427",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:20.069718Z",
     "iopub.status.busy": "2022-05-16T23:18:20.068198Z",
     "iopub.status.idle": "2022-05-16T23:18:20.077629Z",
     "shell.execute_reply": "2022-05-16T23:18:20.078082Z",
     "shell.execute_reply.started": "2022-04-19T22:50:15.58802Z"
    },
    "papermill": {
     "duration": 0.110699,
     "end_time": "2022-05-16T23:18:20.078308",
     "exception": false,
     "start_time": "2022-05-16T23:18:19.967609",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import os\n",
    "iskaggle = os.environ.get('KAGGLE_KERNEL_RUN_TYPE', '')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "48759fa0",
   "metadata": {
    "papermill": {
     "duration": 0.088837,
     "end_time": "2022-05-16T23:18:20.263924",
     "exception": false,
     "start_time": "2022-05-16T23:18:20.175087",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### Using Kaggle data on your own machine"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8b1ca433",
   "metadata": {
    "papermill": {
     "duration": 0.15651,
     "end_time": "2022-05-16T23:18:20.564625",
     "exception": false,
     "start_time": "2022-05-16T23:18:20.408115",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Kaggle limits your weekly time using a GPU machine. The limits are very generous, but you may well still find it's not enough! In that case, you'll want to use your own GPU server, or a cloud server such as Colab, Paperspace Gradient, or SageMaker Studio Lab (all of which have free options). To do so, you'll need to be able to download Kaggle datasets.\n",
    "\n",
    "The easiest way to download Kaggle datasets is to use the Kaggle API. You can install this using `pip` by running this in a notebook cell:\n",
    "\n",
    "    !pip install kaggle\n",
    "\n",
    "You need an API key to use the Kaggle API; to get one, click on your profile picture on the Kaggle website, and choose My Account, then click Create New API Token. This will save a file called *kaggle.json* to your PC. You need to copy this key on your GPU server. To do so, open the file you downloaded, copy the contents, and paste them in the following cell (e.g., `creds = '{\"username\":\"xxx\",\"key\":\"xxx\"}'`):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "6c52e24b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:20.878123Z",
     "iopub.status.busy": "2022-05-16T23:18:20.877241Z",
     "iopub.status.idle": "2022-05-16T23:18:20.879522Z",
     "shell.execute_reply": "2022-05-16T23:18:20.878859Z",
     "shell.execute_reply.started": "2022-04-19T22:50:15.619534Z"
    },
    "papermill": {
     "duration": 0.161714,
     "end_time": "2022-05-16T23:18:20.879673",
     "exception": false,
     "start_time": "2022-05-16T23:18:20.717959",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "creds = ''"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f73d2a77",
   "metadata": {
    "papermill": {
     "duration": 0.091829,
     "end_time": "2022-05-16T23:18:21.075422",
     "exception": false,
     "start_time": "2022-05-16T23:18:20.983593",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Then execute this cell (this only needs to be run once):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "41a7fe66",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:21.273100Z",
     "iopub.status.busy": "2022-05-16T23:18:21.272324Z",
     "iopub.status.idle": "2022-05-16T23:18:21.275167Z",
     "shell.execute_reply": "2022-05-16T23:18:21.274735Z",
     "shell.execute_reply.started": "2022-04-19T22:50:15.625454Z"
    },
    "papermill": {
     "duration": 0.105467,
     "end_time": "2022-05-16T23:18:21.275293",
     "exception": false,
     "start_time": "2022-05-16T23:18:21.169826",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "# for working with paths in Python, I recommend using `pathlib.Path`\n",
    "from pathlib import Path\n",
    "\n",
    "cred_path = Path('~/.kaggle/kaggle.json').expanduser()\n",
    "if not cred_path.exists():\n",
    "    cred_path.parent.mkdir(exist_ok=True)\n",
    "    cred_path.write_text(creds)\n",
    "    cred_path.chmod(0o600)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8cc398d8",
   "metadata": {
    "papermill": {
     "duration": 0.096152,
     "end_time": "2022-05-16T23:18:21.470375",
     "exception": false,
     "start_time": "2022-05-16T23:18:21.374223",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Now you can download datasets from Kaggle."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "f9240dd6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:21.669472Z",
     "iopub.status.busy": "2022-05-16T23:18:21.668423Z",
     "iopub.status.idle": "2022-05-16T23:18:21.671180Z",
     "shell.execute_reply": "2022-05-16T23:18:21.670701Z",
     "shell.execute_reply.started": "2022-04-19T22:50:15.636168Z"
    },
    "papermill": {
     "duration": 0.104227,
     "end_time": "2022-05-16T23:18:21.671310",
     "exception": false,
     "start_time": "2022-05-16T23:18:21.567083",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "path = Path('us-patent-phrase-to-phrase-matching')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d3a9b31d",
   "metadata": {
    "papermill": {
     "duration": 0.097029,
     "end_time": "2022-05-16T23:18:21.865372",
     "exception": false,
     "start_time": "2022-05-16T23:18:21.768343",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "And use the Kaggle API to download the dataset to that path, and extract it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "76cd493d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:22.063705Z",
     "iopub.status.busy": "2022-05-16T23:18:22.062698Z",
     "iopub.status.idle": "2022-05-16T23:18:22.065286Z",
     "shell.execute_reply": "2022-05-16T23:18:22.064712Z",
     "shell.execute_reply.started": "2022-04-19T22:50:15.645742Z"
    },
    "papermill": {
     "duration": 0.105373,
     "end_time": "2022-05-16T23:18:22.065424",
     "exception": false,
     "start_time": "2022-05-16T23:18:21.960051",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "if not iskaggle and not path.exists():\n",
    "    import zipfile,kaggle\n",
    "    kaggle.api.competition_download_cli(str(path))\n",
    "    zipfile.ZipFile(f'{path}.zip').extractall(path)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d234710",
   "metadata": {
    "papermill": {
     "duration": 0.096966,
     "end_time": "2022-05-16T23:18:22.259134",
     "exception": false,
     "start_time": "2022-05-16T23:18:22.162168",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Note that you can easily download notebooks from Kaggle and upload them to other cloud services. So if you're low on Kaggle GPU credits, give this a try!"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "04d27700",
   "metadata": {
    "papermill": {
     "duration": 0.096057,
     "end_time": "2022-05-16T23:18:22.452686",
     "exception": false,
     "start_time": "2022-05-16T23:18:22.356629",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Import and EDA"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "e7d21aca",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:22.653940Z",
     "iopub.status.busy": "2022-05-16T23:18:22.653136Z",
     "iopub.status.idle": "2022-05-16T23:18:31.317348Z",
     "shell.execute_reply": "2022-05-16T23:18:31.316520Z",
     "shell.execute_reply.started": "2022-04-19T22:50:15.653623Z"
    },
    "papermill": {
     "duration": 8.767461,
     "end_time": "2022-05-16T23:18:31.317539",
     "exception": false,
     "start_time": "2022-05-16T23:18:22.550078",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\r\n"
     ]
    }
   ],
   "source": [
    "if iskaggle:\n",
    "    path = Path('../input/us-patent-phrase-to-phrase-matching')\n",
    "    ! pip install -q datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e33c3324",
   "metadata": {
    "papermill": {
     "duration": 0.148707,
     "end_time": "2022-05-16T23:18:31.616056",
     "exception": false,
     "start_time": "2022-05-16T23:18:31.467349",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Documents in NLP datasets are generally in one of two main forms:\n",
    "\n",
    "- **Larger documents**: One text file per document, often organised into one folder per category\n",
    "- **Smaller documents**: One document (or document pair, optionally with metadata) per row in a [CSV file](https://realpython.com/python-csv/).\n",
    "\n",
    "Let's look at our data and see what we've got. In Jupyter you can use any bash/shell command by starting a line with a `!`, and use `{}` to include python variables, like so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "abd6e692",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:31.900273Z",
     "iopub.status.busy": "2022-05-16T23:18:31.899522Z",
     "iopub.status.idle": "2022-05-16T23:18:32.550557Z",
     "shell.execute_reply": "2022-05-16T23:18:32.550032Z",
     "shell.execute_reply.started": "2022-04-19T22:50:24.320172Z"
    },
    "papermill": {
     "duration": 0.789889,
     "end_time": "2022-05-16T23:18:32.550692",
     "exception": false,
     "start_time": "2022-05-16T23:18:31.760803",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "sample_submission.csv  test.csv  train.csv\r\n"
     ]
    }
   ],
   "source": [
    "!ls {path}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "925425fe",
   "metadata": {
    "papermill": {
     "duration": 0.086033,
     "end_time": "2022-05-16T23:18:32.724591",
     "exception": false,
     "start_time": "2022-05-16T23:18:32.638558",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "It looks like this competition uses CSV files. For opening, manipulating, and viewing CSV files, it's generally best to use the Pandas library, which is explained brilliantly in [this book](https://wesmckinney.com/book/) by the lead developer (it's also an excellent introduction to matplotlib and numpy, both of which I use in this notebook). Generally it's imported as the abbreviation `pd`."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "1ccad14f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:32.903526Z",
     "iopub.status.busy": "2022-05-16T23:18:32.902787Z",
     "iopub.status.idle": "2022-05-16T23:18:32.904904Z",
     "shell.execute_reply": "2022-05-16T23:18:32.905273Z",
     "shell.execute_reply.started": "2022-04-19T22:50:25.029375Z"
    },
    "papermill": {
     "duration": 0.094094,
     "end_time": "2022-05-16T23:18:32.905420",
     "exception": false,
     "start_time": "2022-05-16T23:18:32.811326",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b8f0f88b",
   "metadata": {
    "papermill": {
     "duration": 0.088323,
     "end_time": "2022-05-16T23:18:33.079444",
     "exception": false,
     "start_time": "2022-05-16T23:18:32.991121",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Let's set a path to our data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "410bf8a8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:33.257861Z",
     "iopub.status.busy": "2022-05-16T23:18:33.257213Z",
     "iopub.status.idle": "2022-05-16T23:18:33.338741Z",
     "shell.execute_reply": "2022-05-16T23:18:33.338223Z",
     "shell.execute_reply.started": "2022-04-19T22:50:25.036197Z"
    },
    "papermill": {
     "duration": 0.173142,
     "end_time": "2022-05-16T23:18:33.338883",
     "exception": false,
     "start_time": "2022-05-16T23:18:33.165741",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "df = pd.read_csv(path/'train.csv')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "121e2210",
   "metadata": {
    "papermill": {
     "duration": 0.087101,
     "end_time": "2022-05-16T23:18:33.513985",
     "exception": false,
     "start_time": "2022-05-16T23:18:33.426884",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "This creates a [DataFrame](https://pandas.pydata.org/docs/user_guide/10min.html), which is a table of named columns, a bit like a database table. To view the first and last rows, and row count of a DataFrame, just type its name:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "4b298ff1",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:33.700493Z",
     "iopub.status.busy": "2022-05-16T23:18:33.699888Z",
     "iopub.status.idle": "2022-05-16T23:18:33.715083Z",
     "shell.execute_reply": "2022-05-16T23:18:33.715531Z",
     "shell.execute_reply.started": "2022-04-19T22:50:25.122204Z"
    },
    "papermill": {
     "duration": 0.114426,
     "end_time": "2022-05-16T23:18:33.715685",
     "exception": false,
     "start_time": "2022-05-16T23:18:33.601259",
     "status": "completed"
    },
    "tags": []
   },
   "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>id</th>\n",
       "      <th>anchor</th>\n",
       "      <th>target</th>\n",
       "      <th>context</th>\n",
       "      <th>score</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>37d61fd2272659b1</td>\n",
       "      <td>abatement</td>\n",
       "      <td>abatement of pollution</td>\n",
       "      <td>A47</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>7b9652b17b68b7a4</td>\n",
       "      <td>abatement</td>\n",
       "      <td>act of abating</td>\n",
       "      <td>A47</td>\n",
       "      <td>0.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>36d72442aefd8232</td>\n",
       "      <td>abatement</td>\n",
       "      <td>active catalyst</td>\n",
       "      <td>A47</td>\n",
       "      <td>0.25</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>5296b0c19e1ce60e</td>\n",
       "      <td>abatement</td>\n",
       "      <td>eliminating process</td>\n",
       "      <td>A47</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>54c1e3b9184cb5b6</td>\n",
       "      <td>abatement</td>\n",
       "      <td>forest region</td>\n",
       "      <td>A47</td>\n",
       "      <td>0.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36468</th>\n",
       "      <td>8e1386cbefd7f245</td>\n",
       "      <td>wood article</td>\n",
       "      <td>wooden article</td>\n",
       "      <td>B44</td>\n",
       "      <td>1.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36469</th>\n",
       "      <td>42d9e032d1cd3242</td>\n",
       "      <td>wood article</td>\n",
       "      <td>wooden box</td>\n",
       "      <td>B44</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36470</th>\n",
       "      <td>208654ccb9e14fa3</td>\n",
       "      <td>wood article</td>\n",
       "      <td>wooden handle</td>\n",
       "      <td>B44</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36471</th>\n",
       "      <td>756ec035e694722b</td>\n",
       "      <td>wood article</td>\n",
       "      <td>wooden material</td>\n",
       "      <td>B44</td>\n",
       "      <td>0.75</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36472</th>\n",
       "      <td>8d135da0b55b8c88</td>\n",
       "      <td>wood article</td>\n",
       "      <td>wooden substrate</td>\n",
       "      <td>B44</td>\n",
       "      <td>0.50</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>36473 rows × 5 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                     id        anchor                  target context  score\n",
       "0      37d61fd2272659b1     abatement  abatement of pollution     A47   0.50\n",
       "1      7b9652b17b68b7a4     abatement          act of abating     A47   0.75\n",
       "2      36d72442aefd8232     abatement         active catalyst     A47   0.25\n",
       "3      5296b0c19e1ce60e     abatement     eliminating process     A47   0.50\n",
       "4      54c1e3b9184cb5b6     abatement           forest region     A47   0.00\n",
       "...                 ...           ...                     ...     ...    ...\n",
       "36468  8e1386cbefd7f245  wood article          wooden article     B44   1.00\n",
       "36469  42d9e032d1cd3242  wood article              wooden box     B44   0.50\n",
       "36470  208654ccb9e14fa3  wood article           wooden handle     B44   0.50\n",
       "36471  756ec035e694722b  wood article         wooden material     B44   0.75\n",
       "36472  8d135da0b55b8c88  wood article        wooden substrate     B44   0.50\n",
       "\n",
       "[36473 rows x 5 columns]"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "14efc19d",
   "metadata": {
    "papermill": {
     "duration": 0.087104,
     "end_time": "2022-05-16T23:18:33.890429",
     "exception": false,
     "start_time": "2022-05-16T23:18:33.803325",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "It's important to carefully read the [dataset description](https://www.kaggle.com/competitions/us-patent-phrase-to-phrase-matching/data) to understand how each of these columns is used.\n",
    "\n",
    "One of the most useful features of `DataFrame` is the `describe()` method:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "21982274",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:34.098355Z",
     "iopub.status.busy": "2022-05-16T23:18:34.097454Z",
     "iopub.status.idle": "2022-05-16T23:18:34.145806Z",
     "shell.execute_reply": "2022-05-16T23:18:34.146201Z",
     "shell.execute_reply.started": "2022-04-19T22:50:25.149735Z"
    },
    "papermill": {
     "duration": 0.16831,
     "end_time": "2022-05-16T23:18:34.146345",
     "exception": false,
     "start_time": "2022-05-16T23:18:33.978035",
     "status": "completed"
    },
    "tags": []
   },
   "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>id</th>\n",
       "      <th>anchor</th>\n",
       "      <th>target</th>\n",
       "      <th>context</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>36473</td>\n",
       "      <td>36473</td>\n",
       "      <td>36473</td>\n",
       "      <td>36473</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>36473</td>\n",
       "      <td>733</td>\n",
       "      <td>29340</td>\n",
       "      <td>106</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>37d61fd2272659b1</td>\n",
       "      <td>component composite coating</td>\n",
       "      <td>composition</td>\n",
       "      <td>H01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>1</td>\n",
       "      <td>152</td>\n",
       "      <td>24</td>\n",
       "      <td>2186</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      id                       anchor       target context\n",
       "count              36473                        36473        36473   36473\n",
       "unique             36473                          733        29340     106\n",
       "top     37d61fd2272659b1  component composite coating  composition     H01\n",
       "freq                   1                          152           24    2186"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.describe(include='object')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "625b4e8b",
   "metadata": {
    "papermill": {
     "duration": 0.08891,
     "end_time": "2022-05-16T23:18:34.323267",
     "exception": false,
     "start_time": "2022-05-16T23:18:34.234357",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "We can see that in the 36473 rows, there are 733 unique anchors, 106 contexts, and nearly 30000 targets. Some anchors are very common, with \"component composite coating\" for instance appearing 152 times.\n",
    "\n",
    "Earlier, I suggested we could represent the input to the model as something like \"*TEXT1: abatement; TEXT2: eliminating process*\". We'll need to add the context to this too. In Pandas, we just use `+` to concatenate, like so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "d950dfad",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:34.525187Z",
     "iopub.status.busy": "2022-05-16T23:18:34.509818Z",
     "iopub.status.idle": "2022-05-16T23:18:34.535058Z",
     "shell.execute_reply": "2022-05-16T23:18:34.534144Z",
     "shell.execute_reply.started": "2022-04-19T22:50:25.226549Z"
    },
    "papermill": {
     "duration": 0.123242,
     "end_time": "2022-05-16T23:18:34.535176",
     "exception": false,
     "start_time": "2022-05-16T23:18:34.411934",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "df['input'] = 'TEXT1: ' + df.context + '; TEXT2: ' + df.target + '; ANC1: ' + df.anchor"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7e74bcd3",
   "metadata": {
    "papermill": {
     "duration": 0.088484,
     "end_time": "2022-05-16T23:18:34.712258",
     "exception": false,
     "start_time": "2022-05-16T23:18:34.623774",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "We can refer to a column (also known as a *series*) either using regular python \"dotted\" notation, or access it like a dictionary. To get the first few rows, use `head()`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c9bbe94c",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:34.896887Z",
     "iopub.status.busy": "2022-05-16T23:18:34.896259Z",
     "iopub.status.idle": "2022-05-16T23:18:34.899014Z",
     "shell.execute_reply": "2022-05-16T23:18:34.899432Z",
     "shell.execute_reply.started": "2022-04-19T22:50:25.25829Z"
    },
    "papermill": {
     "duration": 0.09874,
     "end_time": "2022-05-16T23:18:34.899569",
     "exception": false,
     "start_time": "2022-05-16T23:18:34.800829",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0    TEXT1: A47; TEXT2: abatement of pollution; ANC...\n",
       "1    TEXT1: A47; TEXT2: act of abating; ANC1: abate...\n",
       "2    TEXT1: A47; TEXT2: active catalyst; ANC1: abat...\n",
       "3    TEXT1: A47; TEXT2: eliminating process; ANC1: ...\n",
       "4    TEXT1: A47; TEXT2: forest region; ANC1: abatement\n",
       "Name: input, dtype: object"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.input.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "528ae2cf",
   "metadata": {
    "papermill": {
     "duration": 0.090197,
     "end_time": "2022-05-16T23:18:35.078246",
     "exception": false,
     "start_time": "2022-05-16T23:18:34.988049",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Tokenization"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ff7d7a2a",
   "metadata": {
    "papermill": {
     "duration": 0.08786,
     "end_time": "2022-05-16T23:18:35.254344",
     "exception": false,
     "start_time": "2022-05-16T23:18:35.166484",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Transformers uses a `Dataset` object for storing a... well a dataset, of course! We can create one like so:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "46fe2b83",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:35.438352Z",
     "iopub.status.busy": "2022-05-16T23:18:35.437675Z",
     "iopub.status.idle": "2022-05-16T23:18:37.535196Z",
     "shell.execute_reply": "2022-05-16T23:18:37.534706Z",
     "shell.execute_reply.started": "2022-04-19T22:50:25.267906Z"
    },
    "papermill": {
     "duration": 2.190631,
     "end_time": "2022-05-16T23:18:37.535330",
     "exception": false,
     "start_time": "2022-05-16T23:18:35.344699",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from datasets import Dataset,DatasetDict\n",
    "\n",
    "ds = Dataset.from_pandas(df)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2a9712b2",
   "metadata": {
    "papermill": {
     "duration": 0.094487,
     "end_time": "2022-05-16T23:18:37.724881",
     "exception": false,
     "start_time": "2022-05-16T23:18:37.630394",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Here's how it's displayed in a notebook:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "26089735",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:37.917875Z",
     "iopub.status.busy": "2022-05-16T23:18:37.916717Z",
     "iopub.status.idle": "2022-05-16T23:18:37.920232Z",
     "shell.execute_reply": "2022-05-16T23:18:37.919832Z",
     "shell.execute_reply.started": "2022-04-19T22:50:27.331754Z"
    },
    "papermill": {
     "duration": 0.097799,
     "end_time": "2022-05-16T23:18:37.920352",
     "exception": false,
     "start_time": "2022-05-16T23:18:37.822553",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Dataset({\n",
       "    features: ['id', 'anchor', 'target', 'context', 'score', 'input'],\n",
       "    num_rows: 36473\n",
       "})"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ds"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "73bff7b1",
   "metadata": {
    "papermill": {
     "duration": 0.090915,
     "end_time": "2022-05-16T23:18:38.100934",
     "exception": false,
     "start_time": "2022-05-16T23:18:38.010019",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "But we can't pass the texts directly into a model. A deep learning model expects numbers as inputs, not English sentences! So we need to do two things:\n",
    "\n",
    "- *Tokenization*: Split each text up into words (or actually, as we'll see, into *tokens*)\n",
    "- *Numericalization*: Convert each word (or token) into a number.\n",
    "\n",
    "The details about how this is done actually depend on the particular model we use. So first we'll need to pick a model. There are thousands of models available, but a reasonable starting point for nearly any NLP problem is to use this (replace \"small\" with \"large\" for a slower but more accurate model, once you've finished exploring):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "94f04956",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:38.291309Z",
     "iopub.status.busy": "2022-05-16T23:18:38.290399Z",
     "iopub.status.idle": "2022-05-16T23:18:38.292739Z",
     "shell.execute_reply": "2022-05-16T23:18:38.292199Z",
     "shell.execute_reply.started": "2022-04-19T22:50:27.345436Z"
    },
    "papermill": {
     "duration": 0.103204,
     "end_time": "2022-05-16T23:18:38.292884",
     "exception": false,
     "start_time": "2022-05-16T23:18:38.189680",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "model_nm = 'microsoft/deberta-v3-small'"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f8a437e9",
   "metadata": {
    "papermill": {
     "duration": 0.08842,
     "end_time": "2022-05-16T23:18:38.471586",
     "exception": false,
     "start_time": "2022-05-16T23:18:38.383166",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "`AutoTokenizer` will create a tokenizer appropriate for a given model:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "9a7191a0",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:38.655835Z",
     "iopub.status.busy": "2022-05-16T23:18:38.655114Z",
     "iopub.status.idle": "2022-05-16T23:18:44.932165Z",
     "shell.execute_reply": "2022-05-16T23:18:44.931194Z",
     "shell.execute_reply.started": "2022-04-19T22:50:27.353989Z"
    },
    "papermill": {
     "duration": 6.371289,
     "end_time": "2022-05-16T23:18:44.932308",
     "exception": false,
     "start_time": "2022-05-16T23:18:38.561019",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ac8ba7b571384940bb57dadacf3cdd90",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/52.0 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "51b3529f03ed4d07aed96e6e41159748",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/578 [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f3d8e12d6a4f44a3a9f5ddc2d88a0036",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/2.35M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Special tokens have been added in the vocabulary, make sure the associated word embeddings are fine-tuned or trained.\n"
     ]
    }
   ],
   "source": [
    "from transformers import AutoModelForSequenceClassification,AutoTokenizer\n",
    "tokz = AutoTokenizer.from_pretrained(model_nm)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bd495e28",
   "metadata": {
    "papermill": {
     "duration": 0.095077,
     "end_time": "2022-05-16T23:18:45.131869",
     "exception": false,
     "start_time": "2022-05-16T23:18:45.036792",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Here's an example of how the tokenizer splits a text into \"tokens\" (which are like words, but can be sub-word pieces, as you see below):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "2c972cb4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:45.320094Z",
     "iopub.status.busy": "2022-05-16T23:18:45.319492Z",
     "iopub.status.idle": "2022-05-16T23:18:45.322225Z",
     "shell.execute_reply": "2022-05-16T23:18:45.322660Z",
     "shell.execute_reply.started": "2022-04-19T22:50:32.889936Z"
    },
    "papermill": {
     "duration": 0.099704,
     "end_time": "2022-05-16T23:18:45.322809",
     "exception": false,
     "start_time": "2022-05-16T23:18:45.223105",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['▁G',\n",
       " \"'\",\n",
       " 'day',\n",
       " '▁folks',\n",
       " ',',\n",
       " '▁I',\n",
       " \"'\",\n",
       " 'm',\n",
       " '▁Jeremy',\n",
       " '▁from',\n",
       " '▁fast',\n",
       " '.',\n",
       " 'ai',\n",
       " '!']"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tokz.tokenize(\"G'day folks, I'm Jeremy from fast.ai!\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e3d6eacc",
   "metadata": {
    "papermill": {
     "duration": 0.100207,
     "end_time": "2022-05-16T23:18:45.513113",
     "exception": false,
     "start_time": "2022-05-16T23:18:45.412906",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Uncommon words will be split into pieces. The start of a new word is represented by `▁`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "a577550d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:45.709111Z",
     "iopub.status.busy": "2022-05-16T23:18:45.708496Z",
     "iopub.status.idle": "2022-05-16T23:18:45.711119Z",
     "shell.execute_reply": "2022-05-16T23:18:45.711568Z",
     "shell.execute_reply.started": "2022-04-19T22:50:32.904456Z"
    },
    "papermill": {
     "duration": 0.098427,
     "end_time": "2022-05-16T23:18:45.711709",
     "exception": false,
     "start_time": "2022-05-16T23:18:45.613282",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['▁A',\n",
       " '▁platypus',\n",
       " '▁is',\n",
       " '▁an',\n",
       " '▁or',\n",
       " 'ni',\n",
       " 'tho',\n",
       " 'rhynch',\n",
       " 'us',\n",
       " '▁an',\n",
       " 'at',\n",
       " 'inus',\n",
       " '.']"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tokz.tokenize(\"A platypus is an ornithorhynchus anatinus.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b98a677",
   "metadata": {
    "papermill": {
     "duration": 0.090466,
     "end_time": "2022-05-16T23:18:45.892899",
     "exception": false,
     "start_time": "2022-05-16T23:18:45.802433",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Here's a simple function which tokenizes our inputs:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "5b4df894",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:46.081044Z",
     "iopub.status.busy": "2022-05-16T23:18:46.079465Z",
     "iopub.status.idle": "2022-05-16T23:18:46.081698Z",
     "shell.execute_reply": "2022-05-16T23:18:46.082102Z",
     "shell.execute_reply.started": "2022-04-19T22:50:32.9125Z"
    },
    "papermill": {
     "duration": 0.098195,
     "end_time": "2022-05-16T23:18:46.082230",
     "exception": false,
     "start_time": "2022-05-16T23:18:45.984035",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def tok_func(x): return tokz(x[\"input\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5fc5b51f",
   "metadata": {
    "papermill": {
     "duration": 0.08984,
     "end_time": "2022-05-16T23:18:46.263955",
     "exception": false,
     "start_time": "2022-05-16T23:18:46.174115",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "To run this quickly in parallel on every row in our dataset, use `map`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "9dcae86d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:46.454144Z",
     "iopub.status.busy": "2022-05-16T23:18:46.453303Z",
     "iopub.status.idle": "2022-05-16T23:18:54.033052Z",
     "shell.execute_reply": "2022-05-16T23:18:54.032576Z",
     "shell.execute_reply.started": "2022-04-19T22:50:32.920244Z"
    },
    "papermill": {
     "duration": 7.678097,
     "end_time": "2022-05-16T23:18:54.033180",
     "exception": false,
     "start_time": "2022-05-16T23:18:46.355083",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "b4303ed564424f61b5b6bdf3397bb22a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/37 [00:00<?, ?ba/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "tok_ds = ds.map(tok_func, batched=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0dd77a16",
   "metadata": {
    "papermill": {
     "duration": 0.094896,
     "end_time": "2022-05-16T23:18:54.222495",
     "exception": false,
     "start_time": "2022-05-16T23:18:54.127599",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "This adds a new item to our dataset called `input_ids`. For instance, here is the input and IDs for the first row of our data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "ef6401c5",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:54.427236Z",
     "iopub.status.busy": "2022-05-16T23:18:54.426628Z",
     "iopub.status.idle": "2022-05-16T23:18:54.429396Z",
     "shell.execute_reply": "2022-05-16T23:18:54.429872Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.025726Z"
    },
    "papermill": {
     "duration": 0.10725,
     "end_time": "2022-05-16T23:18:54.430009",
     "exception": false,
     "start_time": "2022-05-16T23:18:54.322759",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "('TEXT1: A47; TEXT2: abatement of pollution; ANC1: abatement',\n",
       " [1,\n",
       "  54453,\n",
       "  435,\n",
       "  294,\n",
       "  336,\n",
       "  5753,\n",
       "  346,\n",
       "  54453,\n",
       "  445,\n",
       "  294,\n",
       "  47284,\n",
       "  265,\n",
       "  6435,\n",
       "  346,\n",
       "  23702,\n",
       "  435,\n",
       "  294,\n",
       "  47284,\n",
       "  2])"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "row = tok_ds[0]\n",
    "row['input'], row['input_ids']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b5dafa61",
   "metadata": {
    "papermill": {
     "duration": 0.096597,
     "end_time": "2022-05-16T23:18:54.624153",
     "exception": false,
     "start_time": "2022-05-16T23:18:54.527556",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "So, what are those IDs and where do they come from? The secret is that there's a list called `vocab` in the tokenizer which contains a unique integer for every possible token string. We can look them up like this, for instance to find the token for the word \"of\":"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "d15fe108",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:54.834285Z",
     "iopub.status.busy": "2022-05-16T23:18:54.833617Z",
     "iopub.status.idle": "2022-05-16T23:18:54.836600Z",
     "shell.execute_reply": "2022-05-16T23:18:54.837081Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.035118Z"
    },
    "papermill": {
     "duration": 0.114718,
     "end_time": "2022-05-16T23:18:54.837222",
     "exception": false,
     "start_time": "2022-05-16T23:18:54.722504",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "265"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tokz.vocab['▁of']"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "eb58e8bc",
   "metadata": {
    "papermill": {
     "duration": 0.105436,
     "end_time": "2022-05-16T23:18:55.050112",
     "exception": false,
     "start_time": "2022-05-16T23:18:54.944676",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Looking above at our input IDs, we do indeed see that `265` appears as expected.\n",
    "\n",
    "Finally, we need to prepare our labels. Transformers always assumes that your labels has the column name `labels`, but in our dataset it's currently `score`. Therefore, we need to rename it:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "1872e8aa",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:55.247145Z",
     "iopub.status.busy": "2022-05-16T23:18:55.246268Z",
     "iopub.status.idle": "2022-05-16T23:18:55.249927Z",
     "shell.execute_reply": "2022-05-16T23:18:55.249454Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.045156Z"
    },
    "papermill": {
     "duration": 0.101969,
     "end_time": "2022-05-16T23:18:55.250039",
     "exception": false,
     "start_time": "2022-05-16T23:18:55.148070",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "tok_ds = tok_ds.rename_columns({'score':'labels'})"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e458a967",
   "metadata": {
    "papermill": {
     "duration": 0.091676,
     "end_time": "2022-05-16T23:18:55.433347",
     "exception": false,
     "start_time": "2022-05-16T23:18:55.341671",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Now that we've prepared our tokens and labels, we need to create our validation set."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66c7a56c",
   "metadata": {
    "papermill": {
     "duration": 0.091929,
     "end_time": "2022-05-16T23:18:55.616679",
     "exception": false,
     "start_time": "2022-05-16T23:18:55.524750",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Test and validation sets"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "64dd7b30",
   "metadata": {
    "papermill": {
     "duration": 0.105152,
     "end_time": "2022-05-16T23:18:55.814337",
     "exception": false,
     "start_time": "2022-05-16T23:18:55.709185",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "You may have noticed that our directory contained another file:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "53780c56",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:56.029583Z",
     "iopub.status.busy": "2022-05-16T23:18:56.028954Z",
     "iopub.status.idle": "2022-05-16T23:18:56.049673Z",
     "shell.execute_reply": "2022-05-16T23:18:56.050181Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.055398Z"
    },
    "papermill": {
     "duration": 0.130433,
     "end_time": "2022-05-16T23:18:56.050339",
     "exception": false,
     "start_time": "2022-05-16T23:18:55.919906",
     "status": "completed"
    },
    "tags": []
   },
   "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>id</th>\n",
       "      <th>anchor</th>\n",
       "      <th>target</th>\n",
       "      <th>context</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>36</td>\n",
       "      <td>36</td>\n",
       "      <td>36</td>\n",
       "      <td>36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>unique</th>\n",
       "      <td>36</td>\n",
       "      <td>34</td>\n",
       "      <td>36</td>\n",
       "      <td>29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>top</th>\n",
       "      <td>4112d61851461f60</td>\n",
       "      <td>el display</td>\n",
       "      <td>inorganic photoconductor drum</td>\n",
       "      <td>G02</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>freq</th>\n",
       "      <td>1</td>\n",
       "      <td>2</td>\n",
       "      <td>1</td>\n",
       "      <td>3</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                      id      anchor                         target context\n",
       "count                 36          36                             36      36\n",
       "unique                36          34                             36      29\n",
       "top     4112d61851461f60  el display  inorganic photoconductor drum     G02\n",
       "freq                   1           2                              1       3"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "eval_df = pd.read_csv(path/'test.csv')\n",
    "eval_df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b72de683",
   "metadata": {
    "papermill": {
     "duration": 0.093872,
     "end_time": "2022-05-16T23:18:56.237423",
     "exception": false,
     "start_time": "2022-05-16T23:18:56.143551",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "This is the *test set*. Possibly the most important idea in machine learning is that of having separate training, validation, and test data sets."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c80d0906",
   "metadata": {
    "heading_collapsed": true,
    "papermill": {
     "duration": 0.091766,
     "end_time": "2022-05-16T23:18:56.421195",
     "exception": false,
     "start_time": "2022-05-16T23:18:56.329429",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### Validation set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce5af6dc",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.091787,
     "end_time": "2022-05-16T23:18:56.605465",
     "exception": false,
     "start_time": "2022-05-16T23:18:56.513678",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "To explain the motivation, let's start simple, and imagine we're trying to fit a model where the true relationship is this quadratic:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "59f936fe",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:56.807733Z",
     "iopub.status.busy": "2022-05-16T23:18:56.806649Z",
     "iopub.status.idle": "2022-05-16T23:18:56.808617Z",
     "shell.execute_reply": "2022-05-16T23:18:56.809106Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.082981Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.110994,
     "end_time": "2022-05-16T23:18:56.809253",
     "exception": false,
     "start_time": "2022-05-16T23:18:56.698259",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def f(x): return -3*x**2 + 2*x + 20"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9b056d51",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.102842,
     "end_time": "2022-05-16T23:18:57.018924",
     "exception": false,
     "start_time": "2022-05-16T23:18:56.916082",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Unfortunately matplotlib (the most common library for plotting in Python) doesn't come with a way to visualize a function, so we'll write something to do this ourselves:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "f79d3475",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:57.213875Z",
     "iopub.status.busy": "2022-05-16T23:18:57.213034Z",
     "iopub.status.idle": "2022-05-16T23:18:57.214858Z",
     "shell.execute_reply": "2022-05-16T23:18:57.215264Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.088386Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.100055,
     "end_time": "2022-05-16T23:18:57.215397",
     "exception": false,
     "start_time": "2022-05-16T23:18:57.115342",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import numpy as np, matplotlib.pyplot as plt\n",
    "\n",
    "def plot_function(f, min=-2.1, max=2.1, color='r'):\n",
    "    x = np.linspace(min,max, 100)[:,None]\n",
    "    plt.plot(x, f(x), color)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cdd515e0",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.093371,
     "end_time": "2022-05-16T23:18:57.404588",
     "exception": false,
     "start_time": "2022-05-16T23:18:57.311217",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Here's what our function looks like:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "f4718265",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:57.598056Z",
     "iopub.status.busy": "2022-05-16T23:18:57.597509Z",
     "iopub.status.idle": "2022-05-16T23:18:57.789107Z",
     "shell.execute_reply": "2022-05-16T23:18:57.789776Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.097166Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.293081,
     "end_time": "2022-05-16T23:18:57.789928",
     "exception": false,
     "start_time": "2022-05-16T23:18:57.496847",
     "status": "completed"
    },
    "tags": []
   },
   "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": [
    "plot_function(f)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b722c832",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.093043,
     "end_time": "2022-05-16T23:18:57.976558",
     "exception": false,
     "start_time": "2022-05-16T23:18:57.883515",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "For instance, perhaps we've measured the height above ground of an object before and after some event. The measurements will have some random error. We can use numpy's random number generator to simulate that. I like to use `seed` when writing about simulations like this so that I know you'll see the same thing I do:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "db44230d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:58.167367Z",
     "iopub.status.busy": "2022-05-16T23:18:58.166496Z",
     "iopub.status.idle": "2022-05-16T23:18:58.169108Z",
     "shell.execute_reply": "2022-05-16T23:18:58.168668Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.304166Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.099419,
     "end_time": "2022-05-16T23:18:58.169220",
     "exception": false,
     "start_time": "2022-05-16T23:18:58.069801",
     "status": "completed"
    },
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from numpy.random import normal,seed,uniform\n",
    "np.random.seed(42)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "889069b7",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.104043,
     "end_time": "2022-05-16T23:18:58.379851",
     "exception": false,
     "start_time": "2022-05-16T23:18:58.275808",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Here's a function `add_noise` that adds some random variation to an array:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "id": "34c80ce2",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:58.582003Z",
     "iopub.status.busy": "2022-05-16T23:18:58.580339Z",
     "iopub.status.idle": "2022-05-16T23:18:58.582612Z",
     "shell.execute_reply": "2022-05-16T23:18:58.583066Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.310049Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.104517,
     "end_time": "2022-05-16T23:18:58.583206",
     "exception": false,
     "start_time": "2022-05-16T23:18:58.478689",
     "status": "completed"
    },
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def noise(x, scale): return normal(scale=scale, size=x.shape)\n",
    "def add_noise(x, mult, add): return x * (1+noise(x,mult)) + noise(x,add)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9de58b05",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.09537,
     "end_time": "2022-05-16T23:18:58.776554",
     "exception": false,
     "start_time": "2022-05-16T23:18:58.681184",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Let's use it to simulate some measurements evenly distributed over time:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "cd6dc8f3",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:58.981118Z",
     "iopub.status.busy": "2022-05-16T23:18:58.980278Z",
     "iopub.status.idle": "2022-05-16T23:18:59.178183Z",
     "shell.execute_reply": "2022-05-16T23:18:59.178770Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.318907Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.309467,
     "end_time": "2022-05-16T23:18:59.178959",
     "exception": false,
     "start_time": "2022-05-16T23:18:58.869492",
     "status": "completed"
    },
    "tags": []
   },
   "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": [
    "x = np.linspace(-2, 2, num=20)[:,None]\n",
    "y = add_noise(f(x), 0.2, 1.3)\n",
    "plt.scatter(x,y);"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1ee44447",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.095978,
     "end_time": "2022-05-16T23:18:59.370378",
     "exception": false,
     "start_time": "2022-05-16T23:18:59.274400",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Now let's see what happens if we *underfit* or *overfit* these predictions. To do that, we'll create a function that fits a polynomial of some degree (e.g. a line is degree 1, quadratic is degree 2, cubic is degree 3, etc). The details of how this function works don't matter too much so feel free to skip over it if you like!  (PS: if you're not sure about the jargon around polynomials, here's a [great video](https://www.youtube.com/watch?v=ffLLmV4mZwU) which teaches you what you'll need to know.)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "59371ec4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:18:59.563588Z",
     "iopub.status.busy": "2022-05-16T23:18:59.562917Z",
     "iopub.status.idle": "2022-05-16T23:19:00.524919Z",
     "shell.execute_reply": "2022-05-16T23:19:00.523968Z",
     "shell.execute_reply.started": "2022-04-19T22:50:40.539771Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 1.060551,
     "end_time": "2022-05-16T23:19:00.525064",
     "exception": false,
     "start_time": "2022-05-16T23:18:59.464513",
     "status": "completed"
    },
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "from sklearn.pipeline import make_pipeline\n",
    "\n",
    "def plot_poly(degree):\n",
    "    model = make_pipeline(PolynomialFeatures(degree), LinearRegression())\n",
    "    model.fit(x, y)\n",
    "    plt.scatter(x,y)\n",
    "    plot_function(model.predict)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "143c4476",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.095759,
     "end_time": "2022-05-16T23:19:00.717839",
     "exception": false,
     "start_time": "2022-05-16T23:19:00.622080",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "So, what happens if we fit a line (a \"degree 1 polynomial\") to our measurements?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "38869436",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:01.002283Z",
     "iopub.status.busy": "2022-05-16T23:19:00.992285Z",
     "iopub.status.idle": "2022-05-16T23:19:01.176825Z",
     "shell.execute_reply": "2022-05-16T23:19:01.177542Z",
     "shell.execute_reply.started": "2022-04-19T22:50:41.552578Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.365113,
     "end_time": "2022-05-16T23:19:01.177706",
     "exception": false,
     "start_time": "2022-05-16T23:19:00.812593",
     "status": "completed"
    },
    "slideshow": {
     "slide_type": "fragment"
    },
    "tags": []
   },
   "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": [
    "plot_poly(1)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5a792455",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.093894,
     "end_time": "2022-05-16T23:19:01.366741",
     "exception": false,
     "start_time": "2022-05-16T23:19:01.272847",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "As you see, the points on the red line (the line we fitted) aren't very close at all. This is *under-fit* -- there's not enough detail in our function to match our data.\n",
    "\n",
    "And what happens if we fit a degree 10 polynomial to our measurements?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "0946043f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:01.577483Z",
     "iopub.status.busy": "2022-05-16T23:19:01.576630Z",
     "iopub.status.idle": "2022-05-16T23:19:01.740385Z",
     "shell.execute_reply": "2022-05-16T23:19:01.741143Z",
     "shell.execute_reply.started": "2022-04-19T22:50:41.806965Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.280325,
     "end_time": "2022-05-16T23:19:01.741320",
     "exception": false,
     "start_time": "2022-05-16T23:19:01.460995",
     "status": "completed"
    },
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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": [
    "plot_poly(10)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "97381d3d",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.109375,
     "end_time": "2022-05-16T23:19:01.958176",
     "exception": false,
     "start_time": "2022-05-16T23:19:01.848801",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Well now it fits our data better, but it doesn't look like it'll do a great job predicting points other than those we measured -- especially those in earlier or later time periods. This is *over-fit* -- there's too much detail such that the model fits our points, but not the underlying process we really care about.\n",
    "\n",
    "Let's try a degree 2 polynomial (a quadratic), and compare it to our \"true\" function (in blue):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "96e74956",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:02.168390Z",
     "iopub.status.busy": "2022-05-16T23:19:02.167555Z",
     "iopub.status.idle": "2022-05-16T23:19:02.344144Z",
     "shell.execute_reply": "2022-05-16T23:19:02.343715Z",
     "shell.execute_reply.started": "2022-04-19T22:50:41.991988Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.290263,
     "end_time": "2022-05-16T23:19:02.344261",
     "exception": false,
     "start_time": "2022-05-16T23:19:02.053998",
     "status": "completed"
    },
    "slideshow": {
     "slide_type": "fragment"
    },
    "tags": []
   },
   "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": [
    "plot_poly(2)\n",
    "plot_function(f, color='b')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4789f92a",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.096446,
     "end_time": "2022-05-16T23:19:02.537038",
     "exception": false,
     "start_time": "2022-05-16T23:19:02.440592",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "That's not bad at all!\n",
    "\n",
    "So, how do we recognise whether our models are under-fit, over-fit, or \"just right\"? We use a *validation set*. This is a set of data that we \"hold out\" from training -- we don't let our model see it at all. If you use the fastai library, it automatically creates a validation set for you if you don't have one, and will always report metrics (measurements of the accuracy of a model) using the validation set.\n",
    "\n",
    "The validation set is *only* ever used to see how we're doing. It's *never* used as inputs to training the model.\n",
    "\n",
    "Transformers uses a `DatasetDict` for holding your training and validation sets. To create one that contains 25% of our data for the validation set, and 75% for the training set, use `train_test_split`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "b8b1e366",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:02.734895Z",
     "iopub.status.busy": "2022-05-16T23:19:02.734114Z",
     "iopub.status.idle": "2022-05-16T23:19:02.754237Z",
     "shell.execute_reply": "2022-05-16T23:19:02.754678Z",
     "shell.execute_reply.started": "2022-04-19T22:50:42.182674Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.122148,
     "end_time": "2022-05-16T23:19:02.754828",
     "exception": false,
     "start_time": "2022-05-16T23:19:02.632680",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DatasetDict({\n",
       "    train: Dataset({\n",
       "        features: ['id', 'anchor', 'target', 'context', 'labels', 'input', 'input_ids', 'token_type_ids', 'attention_mask'],\n",
       "        num_rows: 27354\n",
       "    })\n",
       "    test: Dataset({\n",
       "        features: ['id', 'anchor', 'target', 'context', 'labels', 'input', 'input_ids', 'token_type_ids', 'attention_mask'],\n",
       "        num_rows: 9119\n",
       "    })\n",
       "})"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dds = tok_ds.train_test_split(0.25, seed=42)\n",
    "dds"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "776f4394",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.096073,
     "end_time": "2022-05-16T23:19:02.948686",
     "exception": false,
     "start_time": "2022-05-16T23:19:02.852613",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "As you see above, the validation set here is called `test` and not `validate`, so be careful!\n",
    "\n",
    "In practice, a random split like we've used here might not be a good idea -- here's what Dr Rachel Thomas has to say about it:\n",
    "\n",
    "> \"*One of the most likely culprits for this disconnect between results in development vs results in production is a poorly chosen validation set (or even worse, no validation set at all). Depending on the nature of your data, choosing a validation set can be the most important step. Although sklearn offers a `train_test_split` method, this method takes a random subset of the data, which is a poor choice for many real-world problems.*\"\n",
    "\n",
    "I strongly recommend reading her article [How (and why) to create a good validation set](https://www.fast.ai/2017/11/13/validation-sets/) to more fully understand this critical topic."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dc1f53c9",
   "metadata": {
    "heading_collapsed": true,
    "papermill": {
     "duration": 0.096584,
     "end_time": "2022-05-16T23:19:03.142638",
     "exception": false,
     "start_time": "2022-05-16T23:19:03.046054",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "### Test set"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e665eb75",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.10414,
     "end_time": "2022-05-16T23:19:03.343573",
     "exception": false,
     "start_time": "2022-05-16T23:19:03.239433",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "So that's the validation set explained, and created. What about the \"test set\" then -- what's that for?\n",
    "\n",
    "The *test set* is yet another dataset that's held out from training. But it's held out from reporting metrics too! The accuracy of your model on the test set is only ever checked after you've completed your entire training process, including trying different models, training methods, data processing, etc.\n",
    "\n",
    "You see, as you try all these different things, to see their impact on the metrics on the validation set, you might just accidentally find a few things that entirely coincidentally improve your validation set metrics, but aren't really better in practice. Given enough time and experiments, you'll find lots of these coincidental improvements. That means you're actually over-fitting to your validation set!\n",
    "\n",
    "That's why we keep a test set held back. Kaggle's public leaderboard is like a test set that you can check from time to time. But don't check too often, or you'll be even over-fitting to the test set!\n",
    "\n",
    "Kaggle has a *second* test set, which is yet another held-out dataset that's only used at the *end* of the competition to assess your predictions. That's called the \"private leaderboard\". Here's a [great post](https://gregpark.io/blog/Kaggle-Psychopathy-Postmortem/) about what can happen if you overfit to the public leaderboard.\n",
    "\n",
    "We'll use `eval` as our name for the test set, to avoid confusion with the `test` dataset that was created above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "3a064b7f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:03.544188Z",
     "iopub.status.busy": "2022-05-16T23:19:03.543651Z",
     "iopub.status.idle": "2022-05-16T23:19:04.825203Z",
     "shell.execute_reply": "2022-05-16T23:19:04.823967Z",
     "shell.execute_reply.started": "2022-04-19T22:50:42.209113Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 1.38502,
     "end_time": "2022-05-16T23:19:04.825395",
     "exception": false,
     "start_time": "2022-05-16T23:19:03.440375",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ca8cd9e5c48c4b17bad4862d247c0565",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/1 [00:00<?, ?ba/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "eval_df['input'] = 'TEXT1: ' + eval_df.context + '; TEXT2: ' + eval_df.target + '; ANC1: ' + eval_df.anchor\n",
    "eval_ds = Dataset.from_pandas(eval_df).map(tok_func, batched=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a782e55e",
   "metadata": {
    "heading_collapsed": true,
    "papermill": {
     "duration": 0.109004,
     "end_time": "2022-05-16T23:19:05.050664",
     "exception": false,
     "start_time": "2022-05-16T23:19:04.941660",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Metrics and correlation"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "afde9fc3",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.097068,
     "end_time": "2022-05-16T23:19:05.244731",
     "exception": false,
     "start_time": "2022-05-16T23:19:05.147663",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "When we're training a model, there will be one or more *metrics* that we're interested in maximising or minimising. These are the measurements that should, hopefully, represent how well our model will works for us.\n",
    "\n",
    "In real life, outside of Kaggle, things not easy... As my partner Dr Rachel Thomas notes in [The problem with metrics is a big problem for AI](https://www.fast.ai/2019/09/24/metrics/):\n",
    "\n",
    ">  At their heart, what most current AI approaches do is to optimize metrics. The practice of optimizing metrics is not new nor unique to AI, yet AI can be particularly efficient (even too efficient!) at doing so. This is important to understand, because any risks of optimizing metrics are heightened by AI. While metrics can be useful in their proper place, there are harms when they are unthinkingly applied. Some of the scariest instances of algorithms run amok all result from over-emphasizing metrics. We have to understand this dynamic in order to understand the urgent risks we are facing due to misuse of AI.\n",
    "\n",
    "In Kaggle, however, it's very straightforward to know what metric to use: Kaggle will tell you! According to this competition's [evaluation page](https://www.kaggle.com/competitions/us-patent-phrase-to-phrase-matching/overview/evaluation), \"*submissions are evaluated on the [Pearson correlation coefficient](https://en.wikipedia.org/wiki/Pearson_correlation_coefficient) between the predicted and actual similarity scores*.\" This coefficient is usually abbreviated using the single letter *r*. It is the most widely used measure of the degree of relationship between two variables.\n",
    "\n",
    "r can vary between `-1`, which means perfect inverse correlation, and `+1`, which means perfect positive correlation. The mathematical formula for it is much less important than getting a good intuition for what the different values look like. To start to get that intuition, let's look at some examples using the [California Housing](https://scikit-learn.org/stable/datasets/real_world.html#california-housing-dataset) dataset, which shows \"*is the median house value for California districts, expressed in hundreds of thousands of dollars*\". This dataset is provided by the excellent [scikit-learn](https://scikit-learn.org/stable/) library, which is the most widely used library for machine learning outside of deep learning."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "0d4d7f9a",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:05.445767Z",
     "iopub.status.busy": "2022-05-16T23:19:05.445083Z",
     "iopub.status.idle": "2022-05-16T23:19:07.853081Z",
     "shell.execute_reply": "2022-05-16T23:19:07.852530Z",
     "shell.execute_reply.started": "2022-04-19T22:50:43.287033Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 2.510087,
     "end_time": "2022-05-16T23:19:07.853218",
     "exception": false,
     "start_time": "2022-05-16T23:19:05.343131",
     "status": "completed"
    },
    "tags": []
   },
   "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>MedInc</th>\n",
       "      <th>HouseAge</th>\n",
       "      <th>AveRooms</th>\n",
       "      <th>AveBedrms</th>\n",
       "      <th>Population</th>\n",
       "      <th>AveOccup</th>\n",
       "      <th>Latitude</th>\n",
       "      <th>Longitude</th>\n",
       "      <th>MedHouseVal</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7506</th>\n",
       "      <td>3.0550</td>\n",
       "      <td>37.0</td>\n",
       "      <td>5.152778</td>\n",
       "      <td>1.048611</td>\n",
       "      <td>729.0</td>\n",
       "      <td>5.062500</td>\n",
       "      <td>33.92</td>\n",
       "      <td>-118.28</td>\n",
       "      <td>1.054</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4720</th>\n",
       "      <td>3.0862</td>\n",
       "      <td>35.0</td>\n",
       "      <td>4.697897</td>\n",
       "      <td>1.055449</td>\n",
       "      <td>1159.0</td>\n",
       "      <td>2.216061</td>\n",
       "      <td>34.05</td>\n",
       "      <td>-118.37</td>\n",
       "      <td>3.453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12888</th>\n",
       "      <td>2.5556</td>\n",
       "      <td>24.0</td>\n",
       "      <td>4.864905</td>\n",
       "      <td>1.129222</td>\n",
       "      <td>1631.0</td>\n",
       "      <td>2.395007</td>\n",
       "      <td>38.66</td>\n",
       "      <td>-121.35</td>\n",
       "      <td>1.057</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13344</th>\n",
       "      <td>3.0057</td>\n",
       "      <td>32.0</td>\n",
       "      <td>4.212687</td>\n",
       "      <td>0.936567</td>\n",
       "      <td>1378.0</td>\n",
       "      <td>5.141791</td>\n",
       "      <td>34.05</td>\n",
       "      <td>-117.64</td>\n",
       "      <td>0.969</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7173</th>\n",
       "      <td>1.9083</td>\n",
       "      <td>42.0</td>\n",
       "      <td>3.888554</td>\n",
       "      <td>1.039157</td>\n",
       "      <td>1535.0</td>\n",
       "      <td>4.623494</td>\n",
       "      <td>34.05</td>\n",
       "      <td>-118.19</td>\n",
       "      <td>1.192</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "       MedInc  HouseAge  AveRooms  AveBedrms  Population  AveOccup  Latitude  \\\n",
       "7506   3.0550      37.0  5.152778   1.048611       729.0  5.062500     33.92   \n",
       "4720   3.0862      35.0  4.697897   1.055449      1159.0  2.216061     34.05   \n",
       "12888  2.5556      24.0  4.864905   1.129222      1631.0  2.395007     38.66   \n",
       "13344  3.0057      32.0  4.212687   0.936567      1378.0  5.141791     34.05   \n",
       "7173   1.9083      42.0  3.888554   1.039157      1535.0  4.623494     34.05   \n",
       "\n",
       "       Longitude  MedHouseVal  \n",
       "7506     -118.28        1.054  \n",
       "4720     -118.37        3.453  \n",
       "12888    -121.35        1.057  \n",
       "13344    -117.64        0.969  \n",
       "7173     -118.19        1.192  "
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from sklearn.datasets import fetch_california_housing\n",
    "housing = fetch_california_housing(as_frame=True)\n",
    "housing = housing['data'].join(housing['target']).sample(1000, random_state=52)\n",
    "housing.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2019b08b",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.107065,
     "end_time": "2022-05-16T23:19:08.070958",
     "exception": false,
     "start_time": "2022-05-16T23:19:07.963893",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "We can see all the correlation coefficients for every combination of columns in this dataset by calling `np.corrcoef`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "3a95a2b7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:08.288467Z",
     "iopub.status.busy": "2022-05-16T23:19:08.283123Z",
     "iopub.status.idle": "2022-05-16T23:19:08.296176Z",
     "shell.execute_reply": "2022-05-16T23:19:08.296946Z",
     "shell.execute_reply.started": "2022-04-19T22:50:45.696764Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.128649,
     "end_time": "2022-05-16T23:19:08.297169",
     "exception": false,
     "start_time": "2022-05-16T23:19:08.168520",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 1.  , -0.12,  0.43, -0.08,  0.01, -0.07, -0.12,  0.04,  0.68],\n",
       "       [-0.12,  1.  , -0.17, -0.06, -0.31,  0.  ,  0.03, -0.13,  0.12],\n",
       "       [ 0.43, -0.17,  1.  ,  0.76, -0.09, -0.07,  0.12, -0.03,  0.21],\n",
       "       [-0.08, -0.06,  0.76,  1.  , -0.08, -0.07,  0.09,  0.  , -0.04],\n",
       "       [ 0.01, -0.31, -0.09, -0.08,  1.  ,  0.16, -0.15,  0.13,  0.  ],\n",
       "       [-0.07,  0.  , -0.07, -0.07,  0.16,  1.  , -0.16,  0.17, -0.27],\n",
       "       [-0.12,  0.03,  0.12,  0.09, -0.15, -0.16,  1.  , -0.93, -0.16],\n",
       "       [ 0.04, -0.13, -0.03,  0.  ,  0.13,  0.17, -0.93,  1.  , -0.03],\n",
       "       [ 0.68,  0.12,  0.21, -0.04,  0.  , -0.27, -0.16, -0.03,  1.  ]])"
      ]
     },
     "execution_count": 39,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.set_printoptions(precision=2, suppress=True)\n",
    "\n",
    "np.corrcoef(housing, rowvar=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "04b9ef46",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.099935,
     "end_time": "2022-05-16T23:19:08.505179",
     "exception": false,
     "start_time": "2022-05-16T23:19:08.405244",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "This works well when we're getting a bunch of values at once, but it's overkill when we want a single coefficient:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "911d424d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:08.710479Z",
     "iopub.status.busy": "2022-05-16T23:19:08.709293Z",
     "iopub.status.idle": "2022-05-16T23:19:08.715181Z",
     "shell.execute_reply": "2022-05-16T23:19:08.714739Z",
     "shell.execute_reply.started": "2022-04-19T22:50:45.70789Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.112399,
     "end_time": "2022-05-16T23:19:08.715315",
     "exception": false,
     "start_time": "2022-05-16T23:19:08.602916",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.  , 0.68],\n",
       "       [0.68, 1.  ]])"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.corrcoef(housing.MedInc, housing.MedHouseVal)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a7f0b3ea",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.098461,
     "end_time": "2022-05-16T23:19:08.933883",
     "exception": false,
     "start_time": "2022-05-16T23:19:08.835422",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Therefore, we'll create this little function to just return the single number we need given a pair of variables:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "8565840d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:09.137395Z",
     "iopub.status.busy": "2022-05-16T23:19:09.136447Z",
     "iopub.status.idle": "2022-05-16T23:19:09.139903Z",
     "shell.execute_reply": "2022-05-16T23:19:09.140273Z",
     "shell.execute_reply.started": "2022-04-19T22:50:45.716828Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.108235,
     "end_time": "2022-05-16T23:19:09.140419",
     "exception": false,
     "start_time": "2022-05-16T23:19:09.032184",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.6760250732906"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def corr(x,y): return np.corrcoef(x,y)[0][1]\n",
    "\n",
    "corr(housing.MedInc, housing.MedHouseVal)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "dfbb3155",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.09923,
     "end_time": "2022-05-16T23:19:09.337993",
     "exception": false,
     "start_time": "2022-05-16T23:19:09.238763",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Now we'll look at a few examples of correlations, using this function (the details of the function don't matter too much):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "fea78fe6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:09.544032Z",
     "iopub.status.busy": "2022-05-16T23:19:09.542325Z",
     "iopub.status.idle": "2022-05-16T23:19:09.544661Z",
     "shell.execute_reply": "2022-05-16T23:19:09.545084Z",
     "shell.execute_reply.started": "2022-04-19T22:50:45.727485Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.106227,
     "end_time": "2022-05-16T23:19:09.545232",
     "exception": false,
     "start_time": "2022-05-16T23:19:09.439005",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def show_corr(df, a, b):\n",
    "    x,y = df[a],df[b]\n",
    "    plt.scatter(x,y, alpha=0.5, s=4)\n",
    "    plt.title(f'{a} vs {b}; r: {corr(x, y):.2f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6e252e37",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.106432,
     "end_time": "2022-05-16T23:19:09.749801",
     "exception": false,
     "start_time": "2022-05-16T23:19:09.643369",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "OK, let's check out the correlation between income and house value:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "c3621405",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:09.962327Z",
     "iopub.status.busy": "2022-05-16T23:19:09.961217Z",
     "iopub.status.idle": "2022-05-16T23:19:10.156235Z",
     "shell.execute_reply": "2022-05-16T23:19:10.155794Z",
     "shell.execute_reply.started": "2022-04-19T22:50:45.735339Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.307424,
     "end_time": "2022-05-16T23:19:10.156356",
     "exception": false,
     "start_time": "2022-05-16T23:19:09.848932",
     "status": "completed"
    },
    "tags": []
   },
   "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": [
    "show_corr(housing, 'MedInc', 'MedHouseVal')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d967187d",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.099654,
     "end_time": "2022-05-16T23:19:10.355784",
     "exception": false,
     "start_time": "2022-05-16T23:19:10.256130",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "So that's what a correlation of 0.68 looks like. It's quite a close relationship, but there's still a lot of variation. (Incidentally, this also shows why looking at your data is so important -- we can see clearly in this plot that house prices above $500,000 seem to have been truncated to that maximum value).\n",
    "\n",
    "Let's take a look at another pair:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "3217ef94",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:10.573146Z",
     "iopub.status.busy": "2022-05-16T23:19:10.572263Z",
     "iopub.status.idle": "2022-05-16T23:19:10.754482Z",
     "shell.execute_reply": "2022-05-16T23:19:10.754042Z",
     "shell.execute_reply.started": "2022-04-19T22:50:45.936734Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.299698,
     "end_time": "2022-05-16T23:19:10.754608",
     "exception": false,
     "start_time": "2022-05-16T23:19:10.454910",
     "status": "completed"
    },
    "tags": []
   },
   "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": [
    "show_corr(housing, 'MedInc', 'AveRooms')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e283f2c4",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.118359,
     "end_time": "2022-05-16T23:19:10.974599",
     "exception": false,
     "start_time": "2022-05-16T23:19:10.856240",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "The relationship looks like it is similarly close to the previous example, but r is much lower than the income vs valuation case. Why is that? The reason is that there are a lot of *outliers* -- values of `AveRooms` well outside the mean.\n",
    "\n",
    "r is very sensitive to outliers. If there's outliers in your data, then the relationship between them will dominate the metric. In this case, the houses with a very high number of rooms don't tend to be that valuable, so it's decreasing r from where it would otherwise be.\n",
    "\n",
    "Let's remove the outliers and try again:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "1d5455b4",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:11.210864Z",
     "iopub.status.busy": "2022-05-16T23:19:11.209971Z",
     "iopub.status.idle": "2022-05-16T23:19:11.388989Z",
     "shell.execute_reply": "2022-05-16T23:19:11.388528Z",
     "shell.execute_reply.started": "2022-04-19T22:50:46.141649Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.29918,
     "end_time": "2022-05-16T23:19:11.389107",
     "exception": false,
     "start_time": "2022-05-16T23:19:11.089927",
     "status": "completed"
    },
    "tags": []
   },
   "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": [
    "subset = housing[housing.AveRooms<15]\n",
    "show_corr(subset, 'MedInc', 'AveRooms')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ddee8bc8",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.161694,
     "end_time": "2022-05-16T23:19:11.723774",
     "exception": false,
     "start_time": "2022-05-16T23:19:11.562080",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "As we expected, now the correlation is very similar to our first comparison.\n",
    "\n",
    "Here's another relationship using `AveRooms` on the subset:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "a40703b7",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:12.065306Z",
     "iopub.status.busy": "2022-05-16T23:19:12.063521Z",
     "iopub.status.idle": "2022-05-16T23:19:12.314930Z",
     "shell.execute_reply": "2022-05-16T23:19:12.315888Z",
     "shell.execute_reply.started": "2022-04-19T22:50:46.33992Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.429853,
     "end_time": "2022-05-16T23:19:12.316088",
     "exception": false,
     "start_time": "2022-05-16T23:19:11.886235",
     "status": "completed"
    },
    "tags": []
   },
   "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": [
    "show_corr(subset, 'MedHouseVal', 'AveRooms')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cba1d702",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.199087,
     "end_time": "2022-05-16T23:19:12.712802",
     "exception": false,
     "start_time": "2022-05-16T23:19:12.513715",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "At this level, with r of 0.34, the relationship is becoming quite weak.\n",
    "\n",
    "Let's look at one more:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "2d6937f6",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:13.004823Z",
     "iopub.status.busy": "2022-05-16T23:19:12.996936Z",
     "iopub.status.idle": "2022-05-16T23:19:13.192851Z",
     "shell.execute_reply": "2022-05-16T23:19:13.192201Z",
     "shell.execute_reply.started": "2022-04-19T22:50:46.526726Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.329812,
     "end_time": "2022-05-16T23:19:13.192993",
     "exception": false,
     "start_time": "2022-05-16T23:19:12.863181",
     "status": "completed"
    },
    "tags": []
   },
   "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": [
    "show_corr(subset, 'HouseAge', 'AveRooms')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ce158a24",
   "metadata": {
    "hidden": true,
    "papermill": {
     "duration": 0.103658,
     "end_time": "2022-05-16T23:19:13.401756",
     "exception": false,
     "start_time": "2022-05-16T23:19:13.298098",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "As you see here, a correlation of -0.2 shows a very weak negative trend.\n",
    "\n",
    "We've seen now examples of a variety of levels of correlation coefficient, so hopefully you're getting a good sense of what this metric means.\n",
    "\n",
    "Transformers expects metrics to be returned as a `dict`, since that way the trainer knows what label to use, so let's create a function to do that:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "a5ff917b",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:13.612237Z",
     "iopub.status.busy": "2022-05-16T23:19:13.611377Z",
     "iopub.status.idle": "2022-05-16T23:19:13.614160Z",
     "shell.execute_reply": "2022-05-16T23:19:13.613658Z",
     "shell.execute_reply.started": "2022-04-19T22:50:46.715707Z"
    },
    "hidden": true,
    "papermill": {
     "duration": 0.109782,
     "end_time": "2022-05-16T23:19:13.614283",
     "exception": false,
     "start_time": "2022-05-16T23:19:13.504501",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "def corr_d(eval_pred): return {'pearson': corr(*eval_pred)}"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ed983c51",
   "metadata": {
    "papermill": {
     "duration": 0.102797,
     "end_time": "2022-05-16T23:19:13.819824",
     "exception": false,
     "start_time": "2022-05-16T23:19:13.717027",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Training"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "352007a6",
   "metadata": {
    "papermill": {
     "duration": 0.118052,
     "end_time": "2022-05-16T23:19:14.047292",
     "exception": false,
     "start_time": "2022-05-16T23:19:13.929240",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## Training our model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7cc0a49a",
   "metadata": {
    "papermill": {
     "duration": 0.102335,
     "end_time": "2022-05-16T23:19:14.253540",
     "exception": false,
     "start_time": "2022-05-16T23:19:14.151205",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "To train a model in Transformers we'll need this:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "811f584f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:14.466565Z",
     "iopub.status.busy": "2022-05-16T23:19:14.465852Z",
     "iopub.status.idle": "2022-05-16T23:19:18.918169Z",
     "shell.execute_reply": "2022-05-16T23:19:18.917243Z",
     "shell.execute_reply.started": "2022-04-19T22:50:46.722181Z"
    },
    "papermill": {
     "duration": 4.55931,
     "end_time": "2022-05-16T23:19:18.918306",
     "exception": false,
     "start_time": "2022-05-16T23:19:14.358996",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "from transformers import TrainingArguments,Trainer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bef40f65",
   "metadata": {
    "papermill": {
     "duration": 0.103363,
     "end_time": "2022-05-16T23:19:19.125535",
     "exception": false,
     "start_time": "2022-05-16T23:19:19.022172",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "We pick a batch size that fits our GPU, and small number of epochs so we can run experiments quickly:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "9a7f73b8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:19.336295Z",
     "iopub.status.busy": "2022-05-16T23:19:19.334700Z",
     "iopub.status.idle": "2022-05-16T23:19:19.336896Z",
     "shell.execute_reply": "2022-05-16T23:19:19.337294Z",
     "shell.execute_reply.started": "2022-04-19T22:50:50.493351Z"
    },
    "papermill": {
     "duration": 0.109762,
     "end_time": "2022-05-16T23:19:19.337450",
     "exception": false,
     "start_time": "2022-05-16T23:19:19.227688",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "bs = 128\n",
    "epochs = 4"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1b9defba",
   "metadata": {
    "papermill": {
     "duration": 0.104127,
     "end_time": "2022-05-16T23:19:19.544960",
     "exception": false,
     "start_time": "2022-05-16T23:19:19.440833",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "The most important hyperparameter is the learning rate. fastai provides a learning rate finder to help you figure this out, but Transformers doesn't, so you'll just have to use trial and error. The idea is to find the largest value you can, but which doesn't result in training failing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "95d56aa8",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:19.756018Z",
     "iopub.status.busy": "2022-05-16T23:19:19.755174Z",
     "iopub.status.idle": "2022-05-16T23:19:19.757522Z",
     "shell.execute_reply": "2022-05-16T23:19:19.756988Z",
     "shell.execute_reply.started": "2022-04-19T22:50:50.499493Z"
    },
    "papermill": {
     "duration": 0.109843,
     "end_time": "2022-05-16T23:19:19.757641",
     "exception": false,
     "start_time": "2022-05-16T23:19:19.647798",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "lr = 8e-5"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "456eba4a",
   "metadata": {
    "papermill": {
     "duration": 0.104473,
     "end_time": "2022-05-16T23:19:19.964659",
     "exception": false,
     "start_time": "2022-05-16T23:19:19.860186",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Transformers uses the `TrainingArguments` class to set up arguments. Don't worry too much about the values we're using here -- they should generally work fine in most cases. It's just the 3 parameters above that you may need to change for different models."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "e009173f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:20.241601Z",
     "iopub.status.busy": "2022-05-16T23:19:20.178470Z",
     "iopub.status.idle": "2022-05-16T23:19:20.247751Z",
     "shell.execute_reply": "2022-05-16T23:19:20.247289Z",
     "shell.execute_reply.started": "2022-04-19T22:50:50.511653Z"
    },
    "papermill": {
     "duration": 0.178531,
     "end_time": "2022-05-16T23:19:20.247877",
     "exception": false,
     "start_time": "2022-05-16T23:19:20.069346",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "args = TrainingArguments('outputs', learning_rate=lr, warmup_ratio=0.1, lr_scheduler_type='cosine', fp16=True,\n",
    "    evaluation_strategy=\"epoch\", per_device_train_batch_size=bs, per_device_eval_batch_size=bs*2,\n",
    "    num_train_epochs=epochs, weight_decay=0.01, report_to='none')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6f09673",
   "metadata": {
    "papermill": {
     "duration": 0.10325,
     "end_time": "2022-05-16T23:19:20.455326",
     "exception": false,
     "start_time": "2022-05-16T23:19:20.352076",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "We can now create our model, and `Trainer`, which is a class which combines the data and model together (just like `Learner` in fastai):"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "c242e16d",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:20.669564Z",
     "iopub.status.busy": "2022-05-16T23:19:20.668967Z",
     "iopub.status.idle": "2022-05-16T23:19:41.632888Z",
     "shell.execute_reply": "2022-05-16T23:19:41.632215Z",
     "shell.execute_reply.started": "2022-04-19T22:50:50.57276Z"
    },
    "papermill": {
     "duration": 21.075395,
     "end_time": "2022-05-16T23:19:41.633043",
     "exception": false,
     "start_time": "2022-05-16T23:19:20.557648",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "06f16c9f0e074c379a708f0fefb12ada",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Downloading:   0%|          | 0.00/273M [00:00<?, ?B/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Some weights of the model checkpoint at microsoft/deberta-v3-small were not used when initializing DebertaV2ForSequenceClassification: ['mask_predictions.dense.weight', 'mask_predictions.dense.bias', 'mask_predictions.LayerNorm.bias', 'lm_predictions.lm_head.dense.weight', 'mask_predictions.LayerNorm.weight', 'mask_predictions.classifier.bias', 'lm_predictions.lm_head.bias', 'lm_predictions.lm_head.LayerNorm.bias', 'mask_predictions.classifier.weight', 'lm_predictions.lm_head.LayerNorm.weight', 'lm_predictions.lm_head.dense.bias']\n",
      "- This IS expected if you are initializing DebertaV2ForSequenceClassification from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\n",
      "- This IS NOT expected if you are initializing DebertaV2ForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\n",
      "Some weights of DebertaV2ForSequenceClassification were not initialized from the model checkpoint at microsoft/deberta-v3-small and are newly initialized: ['pooler.dense.weight', 'classifier.bias', 'pooler.dense.bias', 'classifier.weight']\n",
      "You should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\n",
      "Using amp half precision backend\n"
     ]
    }
   ],
   "source": [
    "model = AutoModelForSequenceClassification.from_pretrained(model_nm, num_labels=1)\n",
    "trainer = Trainer(model, args, train_dataset=dds['train'], eval_dataset=dds['test'],\n",
    "                  tokenizer=tokz, compute_metrics=corr_d)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e88b0728",
   "metadata": {
    "papermill": {
     "duration": 0.103838,
     "end_time": "2022-05-16T23:19:41.843310",
     "exception": false,
     "start_time": "2022-05-16T23:19:41.739472",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "As you see, Transformers spits out lots of warnings. You can safely ignore them.\n",
    "\n",
    "Let's train our model!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "1a8f21fc",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:19:42.058507Z",
     "iopub.status.busy": "2022-05-16T23:19:42.057901Z",
     "iopub.status.idle": "2022-05-16T23:24:41.844317Z",
     "shell.execute_reply": "2022-05-16T23:24:41.843875Z",
     "shell.execute_reply.started": "2022-04-19T22:51:06.370314Z"
    },
    "papermill": {
     "duration": 299.896532,
     "end_time": "2022-05-16T23:24:41.844469",
     "exception": false,
     "start_time": "2022-05-16T23:19:41.947937",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The following columns in the training set  don't have a corresponding argument in `DebertaV2ForSequenceClassification.forward` and have been ignored: context, id, target, anchor, input.\n",
      "/opt/conda/lib/python3.7/site-packages/transformers/optimization.py:309: FutureWarning: This implementation of AdamW is deprecated and will be removed in a future version. Use thePyTorch implementation torch.optim.AdamW instead, or set `no_deprecation_warning=True` to disable this warning\n",
      "  FutureWarning,\n",
      "***** Running training *****\n",
      "  Num examples = 27354\n",
      "  Num Epochs = 4\n",
      "  Instantaneous batch size per device = 128\n",
      "  Total train batch size (w. parallel, distributed & accumulation) = 128\n",
      "  Gradient Accumulation steps = 1\n",
      "  Total optimization steps = 856\n",
      "/opt/conda/lib/python3.7/site-packages/transformers/trainer.py:1410: FutureWarning: Non-finite norm encountered in torch.nn.utils.clip_grad_norm_; continuing anyway. Note that the default behavior will change in a future release to error out if a non-finite total norm is encountered. At that point, setting error_if_nonfinite=false will be required to retain the old behavior.\n",
      "  args.max_grad_norm,\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "    <div>\n",
       "      \n",
       "      <progress value='856' max='856' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
       "      [856/856 04:58, Epoch 4/4]\n",
       "    </div>\n",
       "    <table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       " <tr style=\"text-align: left;\">\n",
       "      <th>Epoch</th>\n",
       "      <th>Training Loss</th>\n",
       "      <th>Validation Loss</th>\n",
       "      <th>Pearson</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>No log</td>\n",
       "      <td>0.024492</td>\n",
       "      <td>0.800443</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>No log</td>\n",
       "      <td>0.022003</td>\n",
       "      <td>0.826113</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>0.041600</td>\n",
       "      <td>0.021423</td>\n",
       "      <td>0.834453</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>0.041600</td>\n",
       "      <td>0.022275</td>\n",
       "      <td>0.834767</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table><p>"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The following columns in the evaluation set  don't have a corresponding argument in `DebertaV2ForSequenceClassification.forward` and have been ignored: context, id, target, anchor, input.\n",
      "***** Running Evaluation *****\n",
      "  Num examples = 9119\n",
      "  Batch size = 256\n",
      "The following columns in the evaluation set  don't have a corresponding argument in `DebertaV2ForSequenceClassification.forward` and have been ignored: context, id, target, anchor, input.\n",
      "***** Running Evaluation *****\n",
      "  Num examples = 9119\n",
      "  Batch size = 256\n",
      "Saving model checkpoint to outputs/checkpoint-500\n",
      "Configuration saved in outputs/checkpoint-500/config.json\n",
      "Model weights saved in outputs/checkpoint-500/pytorch_model.bin\n",
      "tokenizer config file saved in outputs/checkpoint-500/tokenizer_config.json\n",
      "Special tokens file saved in outputs/checkpoint-500/special_tokens_map.json\n",
      "added tokens file saved in outputs/checkpoint-500/added_tokens.json\n",
      "The following columns in the evaluation set  don't have a corresponding argument in `DebertaV2ForSequenceClassification.forward` and have been ignored: context, id, target, anchor, input.\n",
      "***** Running Evaluation *****\n",
      "  Num examples = 9119\n",
      "  Batch size = 256\n",
      "The following columns in the evaluation set  don't have a corresponding argument in `DebertaV2ForSequenceClassification.forward` and have been ignored: context, id, target, anchor, input.\n",
      "***** Running Evaluation *****\n",
      "  Num examples = 9119\n",
      "  Batch size = 256\n",
      "\n",
      "\n",
      "Training completed. Do not forget to share your model on huggingface.co/models =)\n",
      "\n",
      "\n"
     ]
    }
   ],
   "source": [
    "trainer.train();"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f71e4a43",
   "metadata": {
    "papermill": {
     "duration": 0.128839,
     "end_time": "2022-05-16T23:24:42.089146",
     "exception": false,
     "start_time": "2022-05-16T23:24:41.960307",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Lots more warning from Transformers again -- you can ignore these as before.\n",
    "\n",
    "The key thing to look at is the \"Pearson\" value in table above. As you see, it's increasing, and is already above 0.8. That's great news! We can now submit our predictions to Kaggle if we want them to be scored on the official leaderboard. Let's get some predictions on the test set:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "cc9ff8a9",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:24:42.337547Z",
     "iopub.status.busy": "2022-05-16T23:24:42.336666Z",
     "iopub.status.idle": "2022-05-16T23:24:42.387175Z",
     "shell.execute_reply": "2022-05-16T23:24:42.387621Z",
     "shell.execute_reply.started": "2022-04-19T22:56:03.886175Z"
    },
    "papermill": {
     "duration": 0.176198,
     "end_time": "2022-05-16T23:24:42.387780",
     "exception": false,
     "start_time": "2022-05-16T23:24:42.211582",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "The following columns in the test set  don't have a corresponding argument in `DebertaV2ForSequenceClassification.forward` and have been ignored: context, id, target, anchor, input.\n",
      "***** Running Prediction *****\n",
      "  Num examples = 36\n",
      "  Batch size = 256\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "\n",
       "    <div>\n",
       "      \n",
       "      <progress value='1' max='1' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
       "      [1/1 : < :]\n",
       "    </div>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "array([[ 0.51],\n",
       "       [ 0.65],\n",
       "       [ 0.5 ],\n",
       "       [ 0.32],\n",
       "       [-0.04],\n",
       "       [ 0.52],\n",
       "       [ 0.52],\n",
       "       [ 0.07],\n",
       "       [ 0.28],\n",
       "       [ 1.11],\n",
       "       [ 0.25],\n",
       "       [ 0.22],\n",
       "       [ 0.71],\n",
       "       [ 0.88],\n",
       "       [ 0.73],\n",
       "       [ 0.41],\n",
       "       [ 0.33],\n",
       "       [ 0.  ],\n",
       "       [ 0.69],\n",
       "       [ 0.35],\n",
       "       [ 0.4 ],\n",
       "       [ 0.25],\n",
       "       [ 0.12],\n",
       "       [ 0.27],\n",
       "       [ 0.56],\n",
       "       [-0.  ],\n",
       "       [-0.03],\n",
       "       [-0.01],\n",
       "       [-0.03],\n",
       "       [ 0.59],\n",
       "       [ 0.29],\n",
       "       [ 0.03],\n",
       "       [ 0.74],\n",
       "       [ 0.57],\n",
       "       [ 0.46],\n",
       "       [ 0.21]])"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds = trainer.predict(eval_ds).predictions.astype(float)\n",
    "preds"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e7f99fe2",
   "metadata": {
    "papermill": {
     "duration": 0.118083,
     "end_time": "2022-05-16T23:24:42.627541",
     "exception": false,
     "start_time": "2022-05-16T23:24:42.509458",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Look out - some of our predictions are <0, or >1! This once again shows the value of remember to actually *look* at your data. Let's fix those out-of-bounds predictions:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "87e31c26",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:24:42.887528Z",
     "iopub.status.busy": "2022-05-16T23:24:42.886610Z",
     "iopub.status.idle": "2022-05-16T23:24:42.888262Z",
     "shell.execute_reply": "2022-05-16T23:24:42.888836Z",
     "shell.execute_reply.started": "2022-04-19T22:56:03.940986Z"
    },
    "papermill": {
     "duration": 0.130653,
     "end_time": "2022-05-16T23:24:42.888988",
     "exception": false,
     "start_time": "2022-05-16T23:24:42.758335",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "preds = np.clip(preds, 0, 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "73ce77ef",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:24:43.130852Z",
     "iopub.status.busy": "2022-05-16T23:24:43.130226Z",
     "iopub.status.idle": "2022-05-16T23:24:43.133011Z",
     "shell.execute_reply": "2022-05-16T23:24:43.133433Z",
     "shell.execute_reply.started": "2022-04-19T22:56:03.946586Z"
    },
    "papermill": {
     "duration": 0.126485,
     "end_time": "2022-05-16T23:24:43.133567",
     "exception": false,
     "start_time": "2022-05-16T23:24:43.007082",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0.51],\n",
       "       [0.65],\n",
       "       [0.5 ],\n",
       "       [0.32],\n",
       "       [0.  ],\n",
       "       [0.52],\n",
       "       [0.52],\n",
       "       [0.07],\n",
       "       [0.28],\n",
       "       [1.  ],\n",
       "       [0.25],\n",
       "       [0.22],\n",
       "       [0.71],\n",
       "       [0.88],\n",
       "       [0.73],\n",
       "       [0.41],\n",
       "       [0.33],\n",
       "       [0.  ],\n",
       "       [0.69],\n",
       "       [0.35],\n",
       "       [0.4 ],\n",
       "       [0.25],\n",
       "       [0.12],\n",
       "       [0.27],\n",
       "       [0.56],\n",
       "       [0.  ],\n",
       "       [0.  ],\n",
       "       [0.  ],\n",
       "       [0.  ],\n",
       "       [0.59],\n",
       "       [0.29],\n",
       "       [0.03],\n",
       "       [0.74],\n",
       "       [0.57],\n",
       "       [0.46],\n",
       "       [0.21]])"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "preds"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "613fd9ce",
   "metadata": {
    "papermill": {
     "duration": 0.1167,
     "end_time": "2022-05-16T23:24:43.367254",
     "exception": false,
     "start_time": "2022-05-16T23:24:43.250554",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "OK, now we're ready to create our submission file. If you save a CSV in your notebook, you will get the option to submit it later."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "7b89fb1f",
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-05-16T23:24:43.662645Z",
     "iopub.status.busy": "2022-05-16T23:24:43.661829Z",
     "iopub.status.idle": "2022-05-16T23:24:43.714350Z",
     "shell.execute_reply": "2022-05-16T23:24:43.712706Z",
     "shell.execute_reply.started": "2022-04-19T22:56:03.959351Z"
    },
    "papermill": {
     "duration": 0.173814,
     "end_time": "2022-05-16T23:24:43.714480",
     "exception": false,
     "start_time": "2022-05-16T23:24:43.540666",
     "status": "completed"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "69e68b83bfe144bb95d01c779834cf18",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Creating CSV from Arrow format:   0%|          | 0/1 [00:00<?, ?ba/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "857"
      ]
     },
     "execution_count": 58,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import datasets\n",
    "\n",
    "submission = datasets.Dataset.from_dict({\n",
    "    'id': eval_ds['id'],\n",
    "    'score': preds\n",
    "})\n",
    "\n",
    "submission.to_csv('submission.csv', index=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b4b0e3cf",
   "metadata": {
    "papermill": {
     "duration": 0.118094,
     "end_time": "2022-05-16T23:24:43.950959",
     "exception": false,
     "start_time": "2022-05-16T23:24:43.832865",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Unfortunately this is a *code competition* and internet access is disabled. That means the `pip install datasets` command we used above won't work if you want to submit to Kaggle. To fix this, you'll need to download the pip installers to Kaggle first, as [described here](https://www.kaggle.com/c/severstal-steel-defect-detection/discussion/113195). Once you've done that, disable internet in your notebook, go to the Kaggle leaderboards page, and click the *Submission* button."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5cfab1d5",
   "metadata": {
    "papermill": {
     "duration": 0.118314,
     "end_time": "2022-05-16T23:24:44.201057",
     "exception": false,
     "start_time": "2022-05-16T23:24:44.082743",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "## The end"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "59e989c1",
   "metadata": {
    "papermill": {
     "duration": 0.11781,
     "end_time": "2022-05-16T23:24:44.439635",
     "exception": false,
     "start_time": "2022-05-16T23:24:44.321825",
     "status": "completed"
    },
    "tags": []
   },
   "source": [
    "Once you're ready to go deeper, take a look at my [Iterate Like a Grandmaster](https://www.kaggle.com/code/jhoward/iterate-like-a-grandmaster/) notebook.\n",
    "\n",
    "Thanks for reading! This has been a bit of an experiment for me -- I've never done an \"absolute beginners\" guide before on Kaggle. I hope you like it! If you do, I'd greatly appreciate an upvote. Don't hesitate to add a comment if you have any questions or thoughts to add."
   ]
  }
 ],
 "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": "3.7.12"
  },
  "papermill": {
   "default_parameters": {},
   "duration": 396.918297,
   "end_time": "2022-05-16T23:24:47.551396",
   "environment_variables": {},
   "exception": null,
   "input_path": "__notebook__.ipynb",
   "output_path": "__notebook__.ipynb",
   "parameters": {},
   "start_time": "2022-05-16T23:18:10.633099",
   "version": "2.3.3"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "008295fef4124f2187d085515b02b03f": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "06f16c9f0e074c379a708f0fefb12ada": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_49d32794ea3841029c3e495b16ccea97",
        "IPY_MODEL_503ffb73ada44598a4e819536482f7aa",
        "IPY_MODEL_21cf85c5a3cd4a998b46753e9e388dac"
       ],
       "layout": "IPY_MODEL_5f8e3235a83f402e95ed55f226078481"
      }
     },
     "10dc2ddd87e8401493d162303a6d3067": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "159a982bb6c641b2942ce670ba6674f0": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "1c605972a9b74f2b80c607f8eeb9ab86": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "21cf85c5a3cd4a998b46753e9e388dac": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_1c605972a9b74f2b80c607f8eeb9ab86",
       "placeholder": "​",
       "style": "IPY_MODEL_3ad0627d75304941979ab44efaeaa582",
       "value": " 273M/273M [00:12&lt;00:00, 24.5MB/s]"
      }
     },
     "256fc0db05a243189200884fb5615f29": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "2e117ee790ce4f77b007f7343b755656": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "35e8d3e51c034d32a2683117a3681b46": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "36e1ef743c92486297596c43e8087f8b": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "3a264df5e1304dbd86a4a32875f69b6c": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "3ad0627d75304941979ab44efaeaa582": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "3e97062c6a65478ab03c03adc8af248b": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "41a8c06f1c4b471ca7e5f64dee052ed0": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "476fbe79e9c54a3383ab18512ff96cf9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "49d32794ea3841029c3e495b16ccea97": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_9efde56cf8d546539799b23f1b8a4d6e",
       "placeholder": "​",
       "style": "IPY_MODEL_476fbe79e9c54a3383ab18512ff96cf9",
       "value": "Downloading: 100%"
      }
     },
     "4c0f7204d1fe4f25b8f672b7b84dcdf4": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "4d114fb317fc4155b9e1ae54c2b87315": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "4dc6bd3e6cf0414ea68344c03011a309": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_dc33bc5fb61b4efa92c76d639f1af722",
       "max": 578.0,
       "min": 0.0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_fbc96783965c4dcda2e490512b4bab1c",
       "value": 578.0
      }
     },
     "503ffb73ada44598a4e819536482f7aa": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_b56b1e1419754e04832338be487552b0",
       "max": 286059269.0,
       "min": 0.0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_6206d119fd9d44118d1dd9351b74b9af",
       "value": 286059269.0
      }
     },
     "51b3529f03ed4d07aed96e6e41159748": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_985790ebc50e45ef8113f817b1a7bb21",
        "IPY_MODEL_4dc6bd3e6cf0414ea68344c03011a309",
        "IPY_MODEL_663dac2b39f143449c8cfe3ad268b5d9"
       ],
       "layout": "IPY_MODEL_84091969d7f74d0c97670a883ec227e0"
      }
     },
     "5224f65df85f40b79d9cdf9ad421d634": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "537bb7bd2866449a8d9633536d02b38a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "54f390c2a6eb4b91a7f214bbb8ed9f94": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "56d3f6f0e303499a90cfad261dc78989": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "5f8e3235a83f402e95ed55f226078481": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "608f0e67c3dc4b8da44cfe2866f549e0": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_a97fa35881084b82b6db81dbc5baf748",
       "placeholder": "​",
       "style": "IPY_MODEL_f10e1d086c9a469396736baf565b2029",
       "value": "Downloading: 100%"
      }
     },
     "6206d119fd9d44118d1dd9351b74b9af": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "660ff8a0fd564ef8a6ee597dd384eb27": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "663dac2b39f143449c8cfe3ad268b5d9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_3a264df5e1304dbd86a4a32875f69b6c",
       "placeholder": "​",
       "style": "IPY_MODEL_fe83c575183548ac95b73a8941eb1e7d",
       "value": " 578/578 [00:00&lt;00:00, 23.1kB/s]"
      }
     },
     "69e68b83bfe144bb95d01c779834cf18": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_a57e1bd03e714c1d9f9eeadceea5e8c4",
        "IPY_MODEL_c22f1aa9299b44f394352972c055dac3",
        "IPY_MODEL_8ec109e52da042b2b8f21ac037bcd586"
       ],
       "layout": "IPY_MODEL_36e1ef743c92486297596c43e8087f8b"
      }
     },
     "6deff7987a9041fe81c349791054a655": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "6f7b8c8a83d7405b953bcd7d4865b8ff": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_6deff7987a9041fe81c349791054a655",
       "placeholder": "​",
       "style": "IPY_MODEL_b0bb29d13179477aa7e2350bfba39a58",
       "value": "Downloading: 100%"
      }
     },
     "6fd63352ffe744d9a8a338722e5f6c80": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_9743cd05fc074f13831bc705656e10f3",
       "max": 1.0,
       "min": 0.0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_93b8e5ec3e5d401bb8c4c3e752b1f31c",
       "value": 1.0
      }
     },
     "70a35ac44b1346cc8d45771cda5b1b3b": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "749866cf76b945da935d517c6b9cbf9e": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "7ea9429179424f9089fe1ec977f7111c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "80340c65362b48f6b8af7ae4564f651e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_54f390c2a6eb4b91a7f214bbb8ed9f94",
       "placeholder": "​",
       "style": "IPY_MODEL_db5874490df347068c32d5508bc7738a",
       "value": "100%"
      }
     },
     "84091969d7f74d0c97670a883ec227e0": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "88571f1a3d274ebd9ea6bfcbb8234653": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "8b85917f250145d588524063b7ab560f": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "8d636b40e88e415ba36ac264b3885f32": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_4c0f7204d1fe4f25b8f672b7b84dcdf4",
       "placeholder": "​",
       "style": "IPY_MODEL_88571f1a3d274ebd9ea6bfcbb8234653",
       "value": "100%"
      }
     },
     "8ec109e52da042b2b8f21ac037bcd586": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_bc4e82cbac6e4fad8dfeb90e502d3ca2",
       "placeholder": "​",
       "style": "IPY_MODEL_537bb7bd2866449a8d9633536d02b38a",
       "value": " 1/1 [00:00&lt;00:00, 28.34ba/s]"
      }
     },
     "93b8e5ec3e5d401bb8c4c3e752b1f31c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "9743cd05fc074f13831bc705656e10f3": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "9778fcafa45b423abe090310a59efdd7": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "985790ebc50e45ef8113f817b1a7bb21": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_ebb3a108204049d6b8dbc688581c2bba",
       "placeholder": "​",
       "style": "IPY_MODEL_e2fba9146df24492a1cb5a85fdfa3c80",
       "value": "Downloading: 100%"
      }
     },
     "9efde56cf8d546539799b23f1b8a4d6e": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "a57e1bd03e714c1d9f9eeadceea5e8c4": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_749866cf76b945da935d517c6b9cbf9e",
       "placeholder": "​",
       "style": "IPY_MODEL_35e8d3e51c034d32a2683117a3681b46",
       "value": "Creating CSV from Arrow format: 100%"
      }
     },
     "a63ffbf7222749909d590e03c08103d1": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "a97fa35881084b82b6db81dbc5baf748": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "ac8ba7b571384940bb57dadacf3cdd90": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_608f0e67c3dc4b8da44cfe2866f549e0",
        "IPY_MODEL_e8d5be2f3355441ba56c675788d7a863",
        "IPY_MODEL_cc9942b30ad145a7b5c70d177802c3cf"
       ],
       "layout": "IPY_MODEL_10dc2ddd87e8401493d162303a6d3067"
      }
     },
     "af1776634db84292b3ae8e9e0bb0d9a8": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_70a35ac44b1346cc8d45771cda5b1b3b",
       "placeholder": "​",
       "style": "IPY_MODEL_a63ffbf7222749909d590e03c08103d1",
       "value": " 2.35M/2.35M [00:00&lt;00:00, 3.99MB/s]"
      }
     },
     "b0bb29d13179477aa7e2350bfba39a58": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "b4303ed564424f61b5b6bdf3397bb22a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_8d636b40e88e415ba36ac264b3885f32",
        "IPY_MODEL_d3817d49923448d8a969c654b1cd9963",
        "IPY_MODEL_e2a3a37fdef24a19b786d0dfccee779c"
       ],
       "layout": "IPY_MODEL_5224f65df85f40b79d9cdf9ad421d634"
      }
     },
     "b56b1e1419754e04832338be487552b0": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "bc4e82cbac6e4fad8dfeb90e502d3ca2": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "c22f1aa9299b44f394352972c055dac3": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_008295fef4124f2187d085515b02b03f",
       "max": 1.0,
       "min": 0.0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_8b85917f250145d588524063b7ab560f",
       "value": 1.0
      }
     },
     "ca0706bad43c4135acfe3e5c261a8d71": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "ca8cd9e5c48c4b17bad4862d247c0565": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_80340c65362b48f6b8af7ae4564f651e",
        "IPY_MODEL_6fd63352ffe744d9a8a338722e5f6c80",
        "IPY_MODEL_d63928c3e5ae4fd89e655dc5d85659a6"
       ],
       "layout": "IPY_MODEL_2e117ee790ce4f77b007f7343b755656"
      }
     },
     "cb5689148481436bad9d4f548af4458c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "cc9942b30ad145a7b5c70d177802c3cf": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_56d3f6f0e303499a90cfad261dc78989",
       "placeholder": "​",
       "style": "IPY_MODEL_41a8c06f1c4b471ca7e5f64dee052ed0",
       "value": " 52.0/52.0 [00:00&lt;00:00, 1.14kB/s]"
      }
     },
     "cd87e4950c7e4ce0841b1887e28106a1": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "d3817d49923448d8a969c654b1cd9963": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_256fc0db05a243189200884fb5615f29",
       "max": 37.0,
       "min": 0.0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_660ff8a0fd564ef8a6ee597dd384eb27",
       "value": 37.0
      }
     },
     "d63928c3e5ae4fd89e655dc5d85659a6": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_159a982bb6c641b2942ce670ba6674f0",
       "placeholder": "​",
       "style": "IPY_MODEL_3e97062c6a65478ab03c03adc8af248b",
       "value": " 1/1 [00:00&lt;00:00, 18.86ba/s]"
      }
     },
     "db5874490df347068c32d5508bc7738a": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "dc33bc5fb61b4efa92c76d639f1af722": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "e2a3a37fdef24a19b786d0dfccee779c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HTMLModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HTMLModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HTMLView",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_4d114fb317fc4155b9e1ae54c2b87315",
       "placeholder": "​",
       "style": "IPY_MODEL_cb5689148481436bad9d4f548af4458c",
       "value": " 37/37 [00:06&lt;00:00,  4.57ba/s]"
      }
     },
     "e2fba9146df24492a1cb5a85fdfa3c80": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "e472c1bca13b4cf0ac77a3b8f8e0b727": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_ebb78c1af41e4806b16436d1a6e22144",
       "max": 2464616.0,
       "min": 0.0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_7ea9429179424f9089fe1ec977f7111c",
       "value": 2464616.0
      }
     },
     "e8d5be2f3355441ba56c675788d7a863": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "FloatProgressModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "ProgressView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "layout": "IPY_MODEL_ca0706bad43c4135acfe3e5c261a8d71",
       "max": 52.0,
       "min": 0.0,
       "orientation": "horizontal",
       "style": "IPY_MODEL_9778fcafa45b423abe090310a59efdd7",
       "value": 52.0
      }
     },
     "ebb3a108204049d6b8dbc688581c2bba": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "ebb78c1af41e4806b16436d1a6e22144": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "1.2.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_module": "@jupyter-widgets/base",
       "_model_module_version": "1.2.0",
       "_model_name": "LayoutModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "LayoutView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "f10e1d086c9a469396736baf565b2029": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     },
     "f3d8e12d6a4f44a3a9f5ddc2d88a0036": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "HBoxModel",
      "state": {
       "_dom_classes": [],
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "HBoxModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/controls",
       "_view_module_version": "1.5.0",
       "_view_name": "HBoxView",
       "box_style": "",
       "children": [
        "IPY_MODEL_6f7b8c8a83d7405b953bcd7d4865b8ff",
        "IPY_MODEL_e472c1bca13b4cf0ac77a3b8f8e0b727",
        "IPY_MODEL_af1776634db84292b3ae8e9e0bb0d9a8"
       ],
       "layout": "IPY_MODEL_cd87e4950c7e4ce0841b1887e28106a1"
      }
     },
     "fbc96783965c4dcda2e490512b4bab1c": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "ProgressStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "bar_color": null,
       "description_width": ""
      }
     },
     "fe83c575183548ac95b73a8941eb1e7d": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "1.5.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_module_version": "1.5.0",
       "_model_name": "DescriptionStyleModel",
       "_view_count": null,
       "_view_module": "@jupyter-widgets/base",
       "_view_module_version": "1.2.0",
       "_view_name": "StyleView",
       "description_width": ""
      }
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}