examples/Classify_text_with_embeddings.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Tce3stUlHN0L"
      },
      "source": [
        "##### Copyright 2025 Google LLC."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "cellView": "form",
        "id": "tuOe1ymfHZPu"
      },
      "outputs": [],
      "source": [
        "# @title Licensed under the Apache License, Version 2.0 (the \"License\");\n",
        "# you may not use this file except in compliance with the License.\n",
        "# You may obtain a copy of the License at\n",
        "#\n",
        "# https://www.apache.org/licenses/LICENSE-2.0\n",
        "#\n",
        "# Unless required by applicable law or agreed to in writing, software\n",
        "# distributed under the License is distributed on an \"AS IS\" BASIS,\n",
        "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
        "# See the License for the specific language governing permissions and\n",
        "# limitations under the License."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "STuxHh6kk3eL"
      },
      "source": [
        "# Classify text with embeddings\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "wUmTFPw2W_UD"
      },
      "source": [
        "<a target=\"_blank\" href=\"https://colab.research.google.com/github/google-gemini/cookbook/blob/main/examples/Classify_text_with_embeddings.ipynb\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" height=30/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "bhT1u-Pof10V"
      },
      "source": [
        "## Overview\n",
        "\n",
        "In this notebook, you'll learn to use the embeddings produced by the Gemini API to train a model that can classify different types of newsgroup posts based on the topic.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "FXq0ygI3BCdQ"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[?25l   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/149.6 kB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K   \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━\u001b[0m \u001b[32m143.4/149.6 kB\u001b[0m \u001b[31m12.7 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r\u001b[2K   \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m149.6/149.6 kB\u001b[0m \u001b[31m3.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h"
          ]
        }
      ],
      "source": [
        "%pip install -qU \"google-genai>=1.0.0\""
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_mwJYXpElYJc"
      },
      "source": [
        "To run the following cell, your API key must be stored it in a Colab Secret named `GOOGLE_API_KEY`. If you don't already have an API key, or you're not sure how to create a Colab Secret, see the [Authentication](https://github.com/google-gemini/cookbook/blob/main/quickstarts/Authentication.ipynb) quickstart for an example."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "tayrk_A2lZ7A"
      },
      "outputs": [],
      "source": [
        "from google.colab import userdata\n",
        "from google import genai\n",
        "API_KEY = userdata.get('GOOGLE_API_KEY')\n",
        "client = genai.Client(api_key=API_KEY)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C5B9sWq0hNEV"
      },
      "source": [
        "## Dataset\n",
        "\n",
        "The [20 Newsgroups Text Dataset](https://scikit-learn.org/0.19/datasets/twenty_newsgroups.html) contains 18,000 newsgroups posts on 20 topics divided into training and test sets. The split between the training and test datasets are based on messages posted before and after a specific date. For this tutorial, you will be using the subsets of the training and test datasets. You will preprocess and organize the data into Pandas dataframes."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "jDoKis4om-Ea"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['alt.atheism',\n",
              " 'comp.graphics',\n",
              " 'comp.os.ms-windows.misc',\n",
              " 'comp.sys.ibm.pc.hardware',\n",
              " 'comp.sys.mac.hardware',\n",
              " 'comp.windows.x',\n",
              " 'misc.forsale',\n",
              " 'rec.autos',\n",
              " 'rec.motorcycles',\n",
              " 'rec.sport.baseball',\n",
              " 'rec.sport.hockey',\n",
              " 'sci.crypt',\n",
              " 'sci.electronics',\n",
              " 'sci.med',\n",
              " 'sci.space',\n",
              " 'soc.religion.christian',\n",
              " 'talk.politics.guns',\n",
              " 'talk.politics.mideast',\n",
              " 'talk.politics.misc',\n",
              " 'talk.religion.misc']"
            ]
          },
          "execution_count": 4,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "from sklearn.datasets import fetch_20newsgroups\n",
        "newsgroups_train = fetch_20newsgroups(subset='train')\n",
        "newsgroups_test = fetch_20newsgroups(subset='test')\n",
        "\n",
        "# View list of class names for dataset\n",
        "newsgroups_train.target_names"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "hDz9MjkNl_FD"
      },
      "source": [
        "Here is an example of what a data point from the training set looks like."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "FPq-56AimOPX"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Lines: 15\n",
            "\n",
            " I was wondering if anyone out there could enlighten me on this car I saw\n",
            "the other day. It was a 2-door sports car, looked to be from the late 60s/\n",
            "early 70s. It was called a Bricklin. The doors were really small. In addition,\n",
            "the front bumper was separate from the rest of the body. This is \n",
            "all I know. If anyone can tellme a model name, engine specs, years\n",
            "of production, where this car is made, history, or whatever info you\n",
            "have on this funky looking car, please e-mail.\n",
            "\n",
            "Thanks,\n",
            "- IL\n",
            "   ---- brought to you by your neighborhood Lerxst ----\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "\n"
          ]
        }
      ],
      "source": [
        "idx = newsgroups_train.data[0].index('Lines')\n",
        "print(newsgroups_train.data[0][idx:])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "A9-DD7wgCx8j"
      },
      "source": [
        "Now you will begin preprocessing the data for this tutorial. Remove any sensitive information like names, email, or redundant parts of the text like `\"From: \"` and `\"\\nSubject: \"`. Organize the information into a Pandas dataframe so it is more readable."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "urpLwp3UmPF3"
      },
      "outputs": [],
      "source": [
        "import re\n",
        "import pandas as pd\n",
        "\n",
        "def preprocess_newsgroup_data(newsgroup_dataset):\n",
        "  # Apply functions to remove names, emails, and extraneous words from data points in newsgroups.data\n",
        "  newsgroup_dataset.data = [re.sub(r'[\\w\\.-]+@[\\w\\.-]+', '', d) for d in newsgroup_dataset.data] # Remove email\n",
        "  newsgroup_dataset.data = [re.sub(r\"\\([^()]*\\)\", \"\", d) for d in newsgroup_dataset.data] # Remove names\n",
        "  newsgroup_dataset.data = [d.replace(\"From: \", \"\") for d in newsgroup_dataset.data] # Remove \"From: \"\n",
        "  newsgroup_dataset.data = [d.replace(\"\\nSubject: \", \"\") for d in newsgroup_dataset.data] # Remove \"\\nSubject: \"\n",
        "\n",
        "  # Cut off each text entry after 5,000 characters\n",
        "  newsgroup_dataset.data = [d[0:5000] if len(d) > 5000 else d for d in newsgroup_dataset.data]\n",
        "\n",
        "  # Put data points into dataframe\n",
        "  df_processed = pd.DataFrame(newsgroup_dataset.data, columns=['Text'])\n",
        "  df_processed['Label'] = newsgroup_dataset.target\n",
        "  # Match label to target name index\n",
        "  df_processed['Class Name'] = ''\n",
        "  for idx, row in df_processed.iterrows():\n",
        "    df_processed.at[idx, 'Class Name'] = newsgroup_dataset.target_names[row['Label']]\n",
        "\n",
        "  return df_processed"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "JMKddQdNnAOV"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "summary": "{\n  \"name\": \"df_train\",\n  \"rows\": 11314,\n  \"fields\": [\n    {\n      \"column\": \"Text\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 11314,\n        \"samples\": [\n          \" CYCLONE AND TEMPEST?????\\nArticle-I.D.: usenet.1pskav$qtu\\nReply-To:  \\nOrganization: Case Western Reserve University, Cleveland, OH \\nLines: 10\\nNNTP-Posting-Host: thor.ins.cwru.edu\\n\\n\\nCould someone please post any info on these systems.\\n\\nThanks.\\nBoB\\n-- \\n---------------------------------------------------------------------- \\nRobert Novitskey | \\\"Pursuing women is similar to banging one's head\\n  |  against a wall...with less opportunity for reward\\\" \\n---------------------------------------------------------------------- \\n\",\n          \" Re: does dos6 defragment??\\nArticle-I.D.: ux1.ardie.272.734097933\\nOrganization: Department of Plant Pathology\\nLines: 30\\n\\nIn article <>   writes:\\n> \\n>Subject: Re: does dos6 defragment??\\n>Date: Tue, 6 Apr 1993 04:02:54 GMT\\n>In article <>,   writes:\\n>|> Geoffrey S. Elbo writes:\\n>|> \\n>|> >Yes, and it is the fastest defrag I've ever watched.  It did a 170MB \\n>|> >hard disk in 20 minutes.\\n>|> \\n>|> \\tI found the MS defrag looks very much like Norton Speedisk.\\n>|> Is it just a strip-down version of the later?\\n>|> \\n>|> \\tI have both Norton Speedisk and Backup, so I was wondering \\n>|> if I need to install MS Backup?\\n>|> \\n>|> Richard\\n>|> \\n>\\n>Yes, defragger IS come from Norton.\\n>If you have Norton Utility, don't bother.\\n\\n\\n     Don't bother if you have CPBackup or Fastback.  They all offer options \\nnot available in the stripped-down MS version .  Examples - no \\nproprietary format , probably no direct DMA access, and no \\ntape drive!\\n\\n     You NEED MS Defrag if you use doublespace to work on the compressed \\nvolume.\\n\",\n          \" For Sale:  Misc IBM stuff\\nOrganization: The Cellar BBS and public access system\\nLines: 10\\n\\n5.25\\\" Internal Low density disk drive.\\n\\nMonochrome monitor\\n\\n8088 motherboard, built in parallel and serial ports, built in mono and\\ncolor output, 7Mhz.\\n\\nLibertarian, atheist, semi-anarchal Techno-Rat.\\n\\nI define \\n\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Label\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 5,\n        \"min\": 0,\n        \"max\": 19,\n        \"num_unique_values\": 20,\n        \"samples\": [\n          7,\n          17,\n          9\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Class Name\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 20,\n        \"samples\": [\n          \"rec.autos\",\n          \"talk.politics.mideast\",\n          \"rec.sport.baseball\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
              "type": "dataframe",
              "variable_name": "df_train"
            },
            "text/html": [
              "\n",
              "  <div id=\"df-39f62e95-2e56-4250-a439-b25ebe68feb3\" class=\"colab-df-container\">\n",
              "    <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>Text</th>\n",
              "      <th>Label</th>\n",
              "      <th>Class Name</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>WHAT car is this!?\\nNntp-Posting-Host: rac3.w...</td>\n",
              "      <td>7</td>\n",
              "      <td>rec.autos</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>SI Clock Poll - Final Call\\nSummary: Final ca...</td>\n",
              "      <td>4</td>\n",
              "      <td>comp.sys.mac.hardware</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>PB questions...\\nOrganization: Purdue Univers...</td>\n",
              "      <td>4</td>\n",
              "      <td>comp.sys.mac.hardware</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Re: Weitek P9000 ?\\nOrganization: Harris Comp...</td>\n",
              "      <td>1</td>\n",
              "      <td>comp.graphics</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>Re: Shuttle Launch Question\\nOrganization: Sm...</td>\n",
              "      <td>14</td>\n",
              "      <td>sci.space</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-39f62e95-2e56-4250-a439-b25ebe68feb3')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-39f62e95-2e56-4250-a439-b25ebe68feb3 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-39f62e95-2e56-4250-a439-b25ebe68feb3');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-ea015642-53b9-4c18-a7dd-428fb31dde41\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-ea015642-53b9-4c18-a7dd-428fb31dde41')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-ea015642-53b9-4c18-a7dd-428fb31dde41 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "                                                Text  Label  \\\n",
              "0   WHAT car is this!?\\nNntp-Posting-Host: rac3.w...      7   \n",
              "1   SI Clock Poll - Final Call\\nSummary: Final ca...      4   \n",
              "2   PB questions...\\nOrganization: Purdue Univers...      4   \n",
              "3   Re: Weitek P9000 ?\\nOrganization: Harris Comp...      1   \n",
              "4   Re: Shuttle Launch Question\\nOrganization: Sm...     14   \n",
              "\n",
              "              Class Name  \n",
              "0              rec.autos  \n",
              "1  comp.sys.mac.hardware  \n",
              "2  comp.sys.mac.hardware  \n",
              "3          comp.graphics  \n",
              "4              sci.space  "
            ]
          },
          "execution_count": 7,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "# Apply preprocessing function to training and test datasets\n",
        "df_train = preprocess_newsgroup_data(newsgroups_train)\n",
        "df_test = preprocess_newsgroup_data(newsgroups_test)\n",
        "\n",
        "df_train.head()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ogEGbg5XDv-T"
      },
      "source": [
        "Next, you will sample some of the data by taking 100 data points in the training dataset, and dropping a few of the categories to run through this tutorial. Choose the science categories to compare."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "C2N7xXhJohLR"
      },
      "outputs": [],
      "source": [
        "def sample_data(df, num_samples, classes_to_keep):\n",
        "  df = df.groupby('Label', as_index = False).apply(lambda x: x.sample(num_samples)).reset_index(drop=True)\n",
        "\n",
        "  df = df[df['Class Name'].str.contains(classes_to_keep)]\n",
        "\n",
        "  # Reset the encoding of the labels after sampling and dropping certain categories\n",
        "  df['Class Name'] = df['Class Name'].astype('category')\n",
        "  df['Encoded Label'] = df['Class Name'].cat.codes\n",
        "\n",
        "  return df"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "jS2g_ZGupBUb"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "<ipython-input-8-4e962cbdf09d>:2: DeprecationWarning: DataFrameGroupBy.apply operated on the grouping columns. This behavior is deprecated, and in a future version of pandas the grouping columns will be excluded from the operation. Either pass `include_groups=False` to exclude the groupings or explicitly select the grouping columns after groupby to silence this warning.\n",
            "  df = df.groupby('Label', as_index = False).apply(lambda x: x.sample(num_samples)).reset_index(drop=True)\n",
            "<ipython-input-8-4e962cbdf09d>:2: DeprecationWarning: DataFrameGroupBy.apply operated on the grouping columns. This behavior is deprecated, and in a future version of pandas the grouping columns will be excluded from the operation. Either pass `include_groups=False` to exclude the groupings or explicitly select the grouping columns after groupby to silence this warning.\n",
            "  df = df.groupby('Label', as_index = False).apply(lambda x: x.sample(num_samples)).reset_index(drop=True)\n"
          ]
        }
      ],
      "source": [
        "TRAIN_NUM_SAMPLES = 100\n",
        "TEST_NUM_SAMPLES = 25\n",
        "CLASSES_TO_KEEP = 'sci' # Class name should contain 'sci' in it to keep science categories\n",
        "df_train = sample_data(df_train, TRAIN_NUM_SAMPLES, CLASSES_TO_KEEP)\n",
        "df_test = sample_data(df_test, TEST_NUM_SAMPLES, CLASSES_TO_KEEP)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "j04TMPY8rV5q"
      },
      "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>count</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Class Name</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>sci.crypt</th>\n",
              "      <td>100</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>sci.electronics</th>\n",
              "      <td>100</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>sci.med</th>\n",
              "      <td>100</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>sci.space</th>\n",
              "      <td>100</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div><br><label><b>dtype:</b> int64</label>"
            ],
            "text/plain": [
              "Class Name\n",
              "sci.crypt          100\n",
              "sci.electronics    100\n",
              "sci.med            100\n",
              "sci.space          100\n",
              "Name: count, dtype: int64"
            ]
          },
          "execution_count": 10,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_train.value_counts('Class Name')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "qMsnfkVDsJlU"
      },
      "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>count</th>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>Class Name</th>\n",
              "      <th></th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>sci.crypt</th>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>sci.electronics</th>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>sci.med</th>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>sci.space</th>\n",
              "      <td>25</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div><br><label><b>dtype:</b> int64</label>"
            ],
            "text/plain": [
              "Class Name\n",
              "sci.crypt          25\n",
              "sci.electronics    25\n",
              "sci.med            25\n",
              "sci.space          25\n",
              "Name: count, dtype: int64"
            ]
          },
          "execution_count": 11,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_test.value_counts('Class Name')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Kr-WlKzXjYWn"
      },
      "source": [
        "## Create the embeddings\n",
        "\n",
        "In this section, you will see how to generate embeddings for a piece of text using the embeddings from the Gemini API. To learn more about embeddings, visit the [embeddings guide](https://ai.google.dev/docs/embeddings_guide).\n",
        "\n",
        "**NOTE**: Embeddings are computed one at a time, large sample sizes can take a long time!"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yPECMeE2xYA_"
      },
      "source": [
        "### API changes to Embeddings\n",
        "\n",
        "For the recent embeddings model, there is a task type parameter and the optional title (only valid with task_type=`RETRIEVAL_DOCUMENT`).\n",
        "\n",
        "These parameters apply only to the recent embeddings models. The task types are:\n",
        "\n",
        "Task Type | Description\n",
        "---       | ---\n",
        "RETRIEVAL_QUERY\t| Specifies the given text is a query in a search/retrieval setting.\n",
        "RETRIEVAL_DOCUMENT | Specifies the given text is a document in a search/retrieval setting.\n",
        "SEMANTIC_SIMILARITY\t| Specifies the given text will be used for Semantic Textual Similarity (STS).\n",
        "CLASSIFICATION\t| Specifies that the embeddings will be used for classification.\n",
        "CLUSTERING\t| Specifies that the embeddings will be used for clustering."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "MTBGKkPQsotz"
      },
      "outputs": [],
      "source": [
        "from tqdm.auto import tqdm\n",
        "from google.genai import types\n",
        "\n",
        "tqdm.pandas()\n",
        "\n",
        "from google.api_core import retry\n",
        "import numpy as np\n",
        "import math\n",
        "\n",
        "def make_embed_text_fn(model):\n",
        "\n",
        "    @retry.Retry(timeout=300.0)\n",
        "    def embed_fn(texts: list[str]) -> list[list[float]]:\n",
        "        # Set the task_type to CLASSIFICATION and embed the batch of texts\n",
        "        embeddings = client.models.embed_content(\n",
        "            model=model,\n",
        "            contents=texts,\n",
        "            config=types.EmbedContentConfig(task_type=\"CLASSIFICATION\"),\n",
        "        ).embeddings\n",
        "        return np.array([embedding.values for embedding in embeddings])\n",
        "\n",
        "    return embed_fn\n",
        "\n",
        "\n",
        "def create_embeddings(df):\n",
        "    MODEL_ID = \"text-embedding-004\" # @param [\"embedding-001\", \"text-embedding-004\",\"gemini-embedding-exp-03-07\"] {\"allow-input\":true, isTemplate: true}\n",
        "    model = f\"models/{MODEL_ID}\"\n",
        "    embed_fn = make_embed_text_fn(model)\n",
        "\n",
        "    batch_size = 100  # at most 100 requests can be in one batch\n",
        "    all_embeddings = []\n",
        "\n",
        "    # Loop over the texts in chunks of batch_size\n",
        "    for i in tqdm(range(0, len(df), batch_size)):\n",
        "        batch = df[\"Text\"].iloc[i : i + batch_size].tolist()\n",
        "        embeddings = embed_fn(batch)\n",
        "        all_embeddings.extend(embeddings)\n",
        "\n",
        "    df[\"Embeddings\"] = all_embeddings\n",
        "    return df"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "AH0yrHUHtHtw"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "e33d1cb90fec47e4bff622b8d4c0c967",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0/4 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "0071a372f55f41c7ac2d30617e31d331",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0/1 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "df_train = create_embeddings(df_train)\n",
        "df_test = create_embeddings(df_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "6G5TvLlmRjHc"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "summary": "{\n  \"name\": \"df_train\",\n  \"rows\": 400,\n  \"fields\": [\n    {\n      \"column\": \"Text\",\n      \"properties\": {\n        \"dtype\": \"string\",\n        \"num_unique_values\": 400,\n        \"samples\": [\n          \" Need to find information about current trends in diabetes.\\nOrganization: SysLang, Inc.\\nLines: 15\\n\\n\\nI looked for diab in my .newsrc and came up with nuthin. Anyone have\\nany good sources for where I can read? In particular, I'm interested\\nin finding out more about intravenous insulin injection for hepatic\\nvein liver activation. \\n\\nAnything that smells like a pointer would be helpful: newsgroup,\\nmailinglist, etc....\\n\\nMany thanks.\\n\\n-- \\n----------Time flies like the wind. Fruit flies like bananas.------------------\\nSteven W. Orr           uunet!world!steveo\\n----------Everybody repeat after me: \\\"We are all individuals.\\\"-----------------\\n\",\n          \" Re: eye dominance\\nOrganization: University of Colorado, Boulder\\nLines: 38\\nNntp-Posting-Host: beagle.colorado.edu\\n\\n  writes:\\n\\n>[reply to  ]\\n> \\n>>Is there a right-eye dominance  as there is an overall\\n>>right-handedness in the population? I mean do most people require less\\n>>lens corrections for the one eye than the other? If so, what kinds of\\n>>percentages can be attached to this?  Thanks.\\n> \\n>There is an \\\"eyedness\\\" analogous to handedness but it has nothing to do\\n>with refractive error.  To see whether you are right or left eyed, roll\\n>up a sheet of paper into a tube and hold it up to either eye like a\\n>telescope.  The eye that you feel more comfortable putting it up to is\\n>your dominant eye.  Refractive error is often different in the two eyes\\n>but has no correlation with handedness.\\n> \\n>David Nye .  Midelfort Clinic, Eau Claire WI\\n>This is patently absurd; but whoever wishes to become a philosopher\\n>must learn not to be frightened by absurdities. -- Bertrand Russell\\n\\n\\nWhat do you mean \\\"more comfortable putting it up to.\\\"  That seems a bit\\nhard to evaluate.  At least for me it is.  \\n\\nStare straight Point with both hands together and clasp so that only the\\npointer fingers are pointing straight forward to a a spot on the wall about\\neight feet away.  First stare at the spot with both eyes open.  Now\\nclose your left eye.  Now open your left eye.  Now close your right eye.\\nnow open your right eye.\\n\\nIf the image jumped more when you closed your right eye, you are right\\neye dominant.\\n\\nIf the image jumped more when you closed your left eye, you are left eye\\ndominant.\\n\\n\\nDeantha\\n\",\n          \" Re: Clipper considered harmful\\nOrganization: InterCon Systems Corporation - Herndon, VA  USA\\nLines: 14\\nDistribution: world\\nReply-To:  \\nNNTP-Posting-Host: chaos.intercon.com\\nX-Newsreader: InterCon TCP/Connect II 1.1\\n\\n  writes:\\n> Since archiving would be such a powerful tool and so easy to do, why \\n> wouldn't it happen? \\n\\nNot enough magtape.\\n\\nSeriously, do you have any idea how much traffic flows through the US phone \\nsystem in a single day?\\n\\n\\nAmanda Walker\\nInterCon Systems Corporation\\n\\n\\n\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Label\",\n      \"properties\": {\n        \"dtype\": \"number\",\n        \"std\": 1,\n        \"min\": 11,\n        \"max\": 14,\n        \"num_unique_values\": 4,\n        \"samples\": [\n          12,\n          14,\n          11\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Class Name\",\n      \"properties\": {\n        \"dtype\": \"category\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          \"sci.electronics\",\n          \"sci.space\",\n          \"sci.crypt\"\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Encoded Label\",\n      \"properties\": {\n        \"dtype\": \"int8\",\n        \"num_unique_values\": 4,\n        \"samples\": [\n          1,\n          3,\n          0\n        ],\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    },\n    {\n      \"column\": \"Embeddings\",\n      \"properties\": {\n        \"dtype\": \"object\",\n        \"semantic_type\": \"\",\n        \"description\": \"\"\n      }\n    }\n  ]\n}",
              "type": "dataframe",
              "variable_name": "df_train"
            },
            "text/html": [
              "\n",
              "  <div id=\"df-b0d28598-62cd-401f-b562-957f73f99448\" class=\"colab-df-container\">\n",
              "    <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>Text</th>\n",
              "      <th>Label</th>\n",
              "      <th>Class Name</th>\n",
              "      <th>Encoded Label</th>\n",
              "      <th>Embeddings</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>1100</th>\n",
              "      <td>Re: Once tapped, your code is no good any mor...</td>\n",
              "      <td>11</td>\n",
              "      <td>sci.crypt</td>\n",
              "      <td>0</td>\n",
              "      <td>[-0.0049194777, 0.01497897, -0.029226799, 0.03...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1101</th>\n",
              "      <td>Re: Facinating facts: 30 bit serial number, p...</td>\n",
              "      <td>11</td>\n",
              "      <td>sci.crypt</td>\n",
              "      <td>0</td>\n",
              "      <td>[-0.008922124, 0.03260296, -0.031974282, 0.041...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1102</th>\n",
              "      <td>Re: The battle is joined\\nNntp-Posting-Host: ...</td>\n",
              "      <td>11</td>\n",
              "      <td>sci.crypt</td>\n",
              "      <td>0</td>\n",
              "      <td>[-0.008806892, 0.03994155, -0.034680437, 0.029...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1103</th>\n",
              "      <td>Re: Would \"clipper\" make a good cover for oth...</td>\n",
              "      <td>11</td>\n",
              "      <td>sci.crypt</td>\n",
              "      <td>0</td>\n",
              "      <td>[-0.0078087077, 0.028755277, -0.036977194, 0.0...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1104</th>\n",
              "      <td>Cryptography FAQ 09/10 - Other Miscellany\\nOrg...</td>\n",
              "      <td>11</td>\n",
              "      <td>sci.crypt</td>\n",
              "      <td>0</td>\n",
              "      <td>[0.013735601, 0.015993226, -0.05161469, 0.0167...</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>\n",
              "    <div class=\"colab-df-buttons\">\n",
              "\n",
              "  <div class=\"colab-df-container\">\n",
              "    <button class=\"colab-df-convert\" onclick=\"convertToInteractive('df-b0d28598-62cd-401f-b562-957f73f99448')\"\n",
              "            title=\"Convert this dataframe to an interactive table.\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "  <svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\" viewBox=\"0 -960 960 960\">\n",
              "    <path d=\"M120-120v-720h720v720H120Zm60-500h600v-160H180v160Zm220 220h160v-160H400v160Zm0 220h160v-160H400v160ZM180-400h160v-160H180v160Zm440 0h160v-160H620v160ZM180-180h160v-160H180v160Zm440 0h160v-160H620v160Z\"/>\n",
              "  </svg>\n",
              "    </button>\n",
              "\n",
              "  <style>\n",
              "    .colab-df-container {\n",
              "      display:flex;\n",
              "      gap: 12px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert {\n",
              "      background-color: #E8F0FE;\n",
              "      border: none;\n",
              "      border-radius: 50%;\n",
              "      cursor: pointer;\n",
              "      display: none;\n",
              "      fill: #1967D2;\n",
              "      height: 32px;\n",
              "      padding: 0 0 0 0;\n",
              "      width: 32px;\n",
              "    }\n",
              "\n",
              "    .colab-df-convert:hover {\n",
              "      background-color: #E2EBFA;\n",
              "      box-shadow: 0px 1px 2px rgba(60, 64, 67, 0.3), 0px 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "      fill: #174EA6;\n",
              "    }\n",
              "\n",
              "    .colab-df-buttons div {\n",
              "      margin-bottom: 4px;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert {\n",
              "      background-color: #3B4455;\n",
              "      fill: #D2E3FC;\n",
              "    }\n",
              "\n",
              "    [theme=dark] .colab-df-convert:hover {\n",
              "      background-color: #434B5C;\n",
              "      box-shadow: 0px 1px 3px 1px rgba(0, 0, 0, 0.15);\n",
              "      filter: drop-shadow(0px 1px 2px rgba(0, 0, 0, 0.3));\n",
              "      fill: #FFFFFF;\n",
              "    }\n",
              "  </style>\n",
              "\n",
              "    <script>\n",
              "      const buttonEl =\n",
              "        document.querySelector('#df-b0d28598-62cd-401f-b562-957f73f99448 button.colab-df-convert');\n",
              "      buttonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "\n",
              "      async function convertToInteractive(key) {\n",
              "        const element = document.querySelector('#df-b0d28598-62cd-401f-b562-957f73f99448');\n",
              "        const dataTable =\n",
              "          await google.colab.kernel.invokeFunction('convertToInteractive',\n",
              "                                                    [key], {});\n",
              "        if (!dataTable) return;\n",
              "\n",
              "        const docLinkHtml = 'Like what you see? Visit the ' +\n",
              "          '<a target=\"_blank\" href=https://colab.research.google.com/notebooks/data_table.ipynb>data table notebook</a>'\n",
              "          + ' to learn more about interactive tables.';\n",
              "        element.innerHTML = '';\n",
              "        dataTable['output_type'] = 'display_data';\n",
              "        await google.colab.output.renderOutput(dataTable, element);\n",
              "        const docLink = document.createElement('div');\n",
              "        docLink.innerHTML = docLinkHtml;\n",
              "        element.appendChild(docLink);\n",
              "      }\n",
              "    </script>\n",
              "  </div>\n",
              "\n",
              "\n",
              "<div id=\"df-f7730ff5-b326-4fb4-b3e1-793d290b9bc0\">\n",
              "  <button class=\"colab-df-quickchart\" onclick=\"quickchart('df-f7730ff5-b326-4fb4-b3e1-793d290b9bc0')\"\n",
              "            title=\"Suggest charts\"\n",
              "            style=\"display:none;\">\n",
              "\n",
              "<svg xmlns=\"http://www.w3.org/2000/svg\" height=\"24px\"viewBox=\"0 0 24 24\"\n",
              "     width=\"24px\">\n",
              "    <g>\n",
              "        <path d=\"M19 3H5c-1.1 0-2 .9-2 2v14c0 1.1.9 2 2 2h14c1.1 0 2-.9 2-2V5c0-1.1-.9-2-2-2zM9 17H7v-7h2v7zm4 0h-2V7h2v10zm4 0h-2v-4h2v4z\"/>\n",
              "    </g>\n",
              "</svg>\n",
              "  </button>\n",
              "\n",
              "<style>\n",
              "  .colab-df-quickchart {\n",
              "      --bg-color: #E8F0FE;\n",
              "      --fill-color: #1967D2;\n",
              "      --hover-bg-color: #E2EBFA;\n",
              "      --hover-fill-color: #174EA6;\n",
              "      --disabled-fill-color: #AAA;\n",
              "      --disabled-bg-color: #DDD;\n",
              "  }\n",
              "\n",
              "  [theme=dark] .colab-df-quickchart {\n",
              "      --bg-color: #3B4455;\n",
              "      --fill-color: #D2E3FC;\n",
              "      --hover-bg-color: #434B5C;\n",
              "      --hover-fill-color: #FFFFFF;\n",
              "      --disabled-bg-color: #3B4455;\n",
              "      --disabled-fill-color: #666;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart {\n",
              "    background-color: var(--bg-color);\n",
              "    border: none;\n",
              "    border-radius: 50%;\n",
              "    cursor: pointer;\n",
              "    display: none;\n",
              "    fill: var(--fill-color);\n",
              "    height: 32px;\n",
              "    padding: 0;\n",
              "    width: 32px;\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart:hover {\n",
              "    background-color: var(--hover-bg-color);\n",
              "    box-shadow: 0 1px 2px rgba(60, 64, 67, 0.3), 0 1px 3px 1px rgba(60, 64, 67, 0.15);\n",
              "    fill: var(--button-hover-fill-color);\n",
              "  }\n",
              "\n",
              "  .colab-df-quickchart-complete:disabled,\n",
              "  .colab-df-quickchart-complete:disabled:hover {\n",
              "    background-color: var(--disabled-bg-color);\n",
              "    fill: var(--disabled-fill-color);\n",
              "    box-shadow: none;\n",
              "  }\n",
              "\n",
              "  .colab-df-spinner {\n",
              "    border: 2px solid var(--fill-color);\n",
              "    border-color: transparent;\n",
              "    border-bottom-color: var(--fill-color);\n",
              "    animation:\n",
              "      spin 1s steps(1) infinite;\n",
              "  }\n",
              "\n",
              "  @keyframes spin {\n",
              "    0% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "      border-left-color: var(--fill-color);\n",
              "    }\n",
              "    20% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    30% {\n",
              "      border-color: transparent;\n",
              "      border-left-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    40% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-top-color: var(--fill-color);\n",
              "    }\n",
              "    60% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "    }\n",
              "    80% {\n",
              "      border-color: transparent;\n",
              "      border-right-color: var(--fill-color);\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "    90% {\n",
              "      border-color: transparent;\n",
              "      border-bottom-color: var(--fill-color);\n",
              "    }\n",
              "  }\n",
              "</style>\n",
              "\n",
              "  <script>\n",
              "    async function quickchart(key) {\n",
              "      const quickchartButtonEl =\n",
              "        document.querySelector('#' + key + ' button');\n",
              "      quickchartButtonEl.disabled = true;  // To prevent multiple clicks.\n",
              "      quickchartButtonEl.classList.add('colab-df-spinner');\n",
              "      try {\n",
              "        const charts = await google.colab.kernel.invokeFunction(\n",
              "            'suggestCharts', [key], {});\n",
              "      } catch (error) {\n",
              "        console.error('Error during call to suggestCharts:', error);\n",
              "      }\n",
              "      quickchartButtonEl.classList.remove('colab-df-spinner');\n",
              "      quickchartButtonEl.classList.add('colab-df-quickchart-complete');\n",
              "    }\n",
              "    (() => {\n",
              "      let quickchartButtonEl =\n",
              "        document.querySelector('#df-f7730ff5-b326-4fb4-b3e1-793d290b9bc0 button');\n",
              "      quickchartButtonEl.style.display =\n",
              "        google.colab.kernel.accessAllowed ? 'block' : 'none';\n",
              "    })();\n",
              "  </script>\n",
              "</div>\n",
              "\n",
              "    </div>\n",
              "  </div>\n"
            ],
            "text/plain": [
              "                                                   Text  Label Class Name  \\\n",
              "1100   Re: Once tapped, your code is no good any mor...     11  sci.crypt   \n",
              "1101   Re: Facinating facts: 30 bit serial number, p...     11  sci.crypt   \n",
              "1102   Re: The battle is joined\\nNntp-Posting-Host: ...     11  sci.crypt   \n",
              "1103   Re: Would \"clipper\" make a good cover for oth...     11  sci.crypt   \n",
              "1104  Cryptography FAQ 09/10 - Other Miscellany\\nOrg...     11  sci.crypt   \n",
              "\n",
              "      Encoded Label                                         Embeddings  \n",
              "1100              0  [-0.0049194777, 0.01497897, -0.029226799, 0.03...  \n",
              "1101              0  [-0.008922124, 0.03260296, -0.031974282, 0.041...  \n",
              "1102              0  [-0.008806892, 0.03994155, -0.034680437, 0.029...  \n",
              "1103              0  [-0.0078087077, 0.028755277, -0.036977194, 0.0...  \n",
              "1104              0  [0.013735601, 0.015993226, -0.05161469, 0.0167...  "
            ]
          },
          "execution_count": 15,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "df_train.head()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "QPYEYkIsWt_5"
      },
      "source": [
        "## Build a simple classification model\n",
        "Here you will define a simple model with one hidden layer and a single class probability output. The prediction will correspond to the probability of a piece of text being a particular class of news. When you build your model, Keras will automatically shuffle the data points."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "3oLGi4w5JsQR"
      },
      "outputs": [],
      "source": [
        "import keras\n",
        "from keras import layers\n",
        "\n",
        "def build_classification_model(input_size: int, num_classes: int) -> keras.Model:\n",
        "  inputs = x = keras.Input(shape=(input_size,))\n",
        "  x = layers.Dense(input_size, activation='relu')(x)\n",
        "  x = layers.Dense(num_classes, activation='softmax')(x)\n",
        "  return keras.Model(inputs=inputs, outputs=x)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "id": "kORA1Akl5GsG"
      },
      "outputs": [
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"functional\"</span>\n",
              "</pre>\n"
            ],
            "text/plain": [
              "\u001b[1mModel: \"functional\"\u001b[0m\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
              "┃<span style=\"font-weight: bold\"> Layer (type)                         </span>┃<span style=\"font-weight: bold\"> Output Shape                </span>┃<span style=\"font-weight: bold\">         Param # </span>┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">InputLayer</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">768</span>)                 │               <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                        │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">768</span>)                 │         <span style=\"color: #00af00; text-decoration-color: #00af00\">590,592</span> │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                      │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">4</span>)                   │           <span style=\"color: #00af00; text-decoration-color: #00af00\">3,076</span> │\n",
              "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n",
              "</pre>\n"
            ],
            "text/plain": [
              "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━┓\n",
              "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                        \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape               \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m        Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
              "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━┩\n",
              "│ input_layer (\u001b[38;5;33mInputLayer\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m768\u001b[0m)                 │               \u001b[38;5;34m0\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense (\u001b[38;5;33mDense\u001b[0m)                        │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m768\u001b[0m)                 │         \u001b[38;5;34m590,592\u001b[0m │\n",
              "├──────────────────────────────────────┼─────────────────────────────┼─────────────────┤\n",
              "│ dense_1 (\u001b[38;5;33mDense\u001b[0m)                      │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m4\u001b[0m)                   │           \u001b[38;5;34m3,076\u001b[0m │\n",
              "└──────────────────────────────────────┴─────────────────────────────┴─────────────────┘\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">593,668</span> (2.26 MB)\n",
              "</pre>\n"
            ],
            "text/plain": [
              "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m593,668\u001b[0m (2.26 MB)\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">593,668</span> (2.26 MB)\n",
              "</pre>\n"
            ],
            "text/plain": [
              "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m593,668\u001b[0m (2.26 MB)\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/html": [
              "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
              "</pre>\n"
            ],
            "text/plain": [
              "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "# Derive the embedding size from the first training element.\n",
        "embedding_size = len(df_train['Embeddings'].iloc[0])\n",
        "\n",
        "# Give your model a different name, as you have already used the variable name 'model'\n",
        "classifier = build_classification_model(embedding_size, len(df_train['Class Name'].unique()))\n",
        "classifier.summary()\n",
        "\n",
        "classifier.compile(loss = keras.losses.SparseCategoricalCrossentropy(from_logits=False),\n",
        "                   optimizer = keras.optimizers.Adam(learning_rate=0.001),\n",
        "                   metrics=['accuracy'])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "iPYYKnqFvt9x"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "768"
            ]
          },
          "execution_count": 18,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "embedding_size"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kbpTGGiMXDxl"
      },
      "source": [
        "## Train the model to classify newsgroups\n",
        "\n",
        "Finally, you can train a simple model. Use a small number of epochs to avoid overfitting. The first epoch takes much longer than the rest, because the embeddings need to be computed only once."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "id": "bGgvMZGfJ1A4"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 30ms/step - accuracy: 0.3106 - loss: 1.3513 - val_accuracy: 0.4200 - val_loss: 1.2723\n",
            "Epoch 2/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.5781 - loss: 1.1828 - val_accuracy: 0.7400 - val_loss: 1.1253\n",
            "Epoch 3/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.8666 - loss: 0.9892 - val_accuracy: 0.8600 - val_loss: 0.9653\n",
            "Epoch 4/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.8673 - loss: 0.8209 - val_accuracy: 0.7800 - val_loss: 0.8620\n",
            "Epoch 5/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - accuracy: 0.9197 - loss: 0.6397 - val_accuracy: 0.8000 - val_loss: 0.7036\n",
            "Epoch 6/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 19ms/step - accuracy: 0.9337 - loss: 0.5040 - val_accuracy: 0.9100 - val_loss: 0.5824\n",
            "Epoch 7/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 14ms/step - accuracy: 0.9598 - loss: 0.3877 - val_accuracy: 0.9000 - val_loss: 0.5111\n",
            "Epoch 8/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.9432 - loss: 0.3187 - val_accuracy: 0.8600 - val_loss: 0.4583\n",
            "Epoch 9/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.9672 - loss: 0.2606 - val_accuracy: 0.8900 - val_loss: 0.4084\n",
            "Epoch 10/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 18ms/step - accuracy: 0.9572 - loss: 0.2243 - val_accuracy: 0.8900 - val_loss: 0.3839\n",
            "Epoch 11/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.9794 - loss: 0.1908 - val_accuracy: 0.8900 - val_loss: 0.3538\n",
            "Epoch 12/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 23ms/step - accuracy: 0.9873 - loss: 0.1607 - val_accuracy: 0.8900 - val_loss: 0.3328\n",
            "Epoch 13/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 24ms/step - accuracy: 0.9890 - loss: 0.1396 - val_accuracy: 0.9100 - val_loss: 0.3446\n",
            "Epoch 14/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 25ms/step - accuracy: 0.9852 - loss: 0.1344 - val_accuracy: 0.9000 - val_loss: 0.3090\n",
            "Epoch 15/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 22ms/step - accuracy: 0.9868 - loss: 0.1121 - val_accuracy: 0.8900 - val_loss: 0.3090\n",
            "Epoch 16/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.9933 - loss: 0.1017 - val_accuracy: 0.9000 - val_loss: 0.2867\n",
            "Epoch 17/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step - accuracy: 0.9996 - loss: 0.0780 - val_accuracy: 0.9100 - val_loss: 0.3012\n",
            "Epoch 18/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 16ms/step - accuracy: 0.9968 - loss: 0.0844 - val_accuracy: 0.9100 - val_loss: 0.2789\n",
            "Epoch 19/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.9995 - loss: 0.0695 - val_accuracy: 0.9100 - val_loss: 0.2714\n",
            "Epoch 20/25\n",
            "\u001b[1m13/13\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 15ms/step - accuracy: 0.9961 - loss: 0.0608 - val_accuracy: 0.8800 - val_loss: 0.2724\n"
          ]
        }
      ],
      "source": [
        "NUM_EPOCHS = 25\n",
        "BATCH_SIZE = 32\n",
        "\n",
        "# Split the x and y components of the train and validation subsets.\n",
        "y_train = df_train['Encoded Label']\n",
        "x_train = np.stack(df_train['Embeddings'])\n",
        "y_val = df_test['Encoded Label']\n",
        "x_val = np.stack(df_test['Embeddings'])\n",
        "\n",
        "# Train the model for the desired number of epochs.\n",
        "callback = keras.callbacks.EarlyStopping(monitor='accuracy', patience=3)\n",
        "\n",
        "history = classifier.fit(x=x_train,\n",
        "                         y=y_train,\n",
        "                         validation_data=(x_val, y_val),\n",
        "                         callbacks=[callback],\n",
        "                         batch_size=BATCH_SIZE,\n",
        "                         epochs=NUM_EPOCHS,)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "xGBaDHZUPdJO"
      },
      "source": [
        "## Evaluate model performance\n",
        "\n",
        "Use Keras <a href=\"https://www.tensorflow.org/api_docs/python/tf/keras/Model#evaluate\"><code>Model.evaluate</code></a> to get the loss and accuracy on the test dataset."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "id": "d2kOeiqqQIB8"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 10ms/step - accuracy: 0.8801 - loss: 0.2847\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "{'accuracy': 0.8799999952316284, 'loss': 0.27236178517341614}"
            ]
          },
          "execution_count": 20,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "classifier.evaluate(x=x_val, y=y_val, return_dict=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "UyxMhiLYQXAN"
      },
      "source": [
        "One way to evaluate your model performance is to visualize the classifier performance. Use `plot_history` to see the loss and accuracy trends over the epochs."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "MaDO9hwbEOW3"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 2000x800 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "def plot_history(history):\n",
        "  \"\"\"\n",
        "    Plotting training and validation learning curves.\n",
        "\n",
        "    Args:\n",
        "      history: model history with all the metric measures\n",
        "  \"\"\"\n",
        "  fig, (ax1, ax2) = plt.subplots(1,2)\n",
        "  fig.set_size_inches(20, 8)\n",
        "\n",
        "  # Plot loss\n",
        "  ax1.set_title('Loss')\n",
        "  ax1.plot(history.history['loss'], label = 'train')\n",
        "  ax1.plot(history.history['val_loss'], label = 'test')\n",
        "  ax1.set_ylabel('Loss')\n",
        "\n",
        "  ax1.set_xlabel('Epoch')\n",
        "  ax1.legend(['Train', 'Validation'])\n",
        "\n",
        "  # Plot accuracy\n",
        "  ax2.set_title('Accuracy')\n",
        "  ax2.plot(history.history['accuracy'],  label = 'train')\n",
        "  ax2.plot(history.history['val_accuracy'], label = 'test')\n",
        "  ax2.set_ylabel('Accuracy')\n",
        "  ax2.set_xlabel('Epoch')\n",
        "  ax2.legend(['Train', 'Validation'])\n",
        "\n",
        "  plt.show()\n",
        "\n",
        "plot_history(history)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kOgva0pbP4FS"
      },
      "source": [
        "Another way to view model performance, beyond just measuring loss and accuracy is to use a confusion matrix. The confusion matrix allows you to assess the performance of the classification model beyond accuracy. You can see what misclassified points get classified as. In order to build the confusion matrix for this multi-class classification problem, get the actual values in the test set and the predicted values.\n",
        "\n",
        "Start by generating the predicted class for each example in the validation set using [`Model.predict()`](https://www.tensorflow.org/api_docs/python/tf/keras/Model#predict)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "PRUx5ao9QRcO"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "\u001b[1m4/4\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 17ms/step\n"
          ]
        }
      ],
      "source": [
        "y_hat = classifier.predict(x=x_val)\n",
        "y_hat = np.argmax(y_hat, axis=1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "CVidbr0OT5tL"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{'sci.crypt': 0, 'sci.electronics': 1, 'sci.med': 2, 'sci.space': 3}"
            ]
          },
          "execution_count": 23,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "labels_dict = dict(zip(df_test['Class Name'], df_test['Encoded Label']))\n",
        "labels_dict"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "id": "3ae76701e178"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "import sklearn.metrics as skmetrics\n",
        "\n",
        "cm = skmetrics.confusion_matrix(y_val, y_hat)\n",
        "disp = skmetrics.ConfusionMatrixDisplay(confusion_matrix=cm,\n",
        "                              display_labels=labels_dict.keys())\n",
        "disp.plot(xticks_rotation='vertical')\n",
        "plt.title('Confusion matrix for newsgroup test dataset');\n",
        "plt.grid(False)"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "name": "Classify_text_with_embeddings.ipynb",
      "toc_visible": true
    },
    "google": {
      "image_path": "/examples/train_text_classifier_embeddings_files/output_3ae76701e178_0.png",
      "keywords": [
        "examples",
        "googleai",
        "samplecode",
        "python",
        "embed"
      ]
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}