anwser_extracter_from_2_sentence.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "anwser_extracter_from_2_sentence.ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "D-HcJgSsPC6r",
        "outputId": "88e1442c-e1f4-4363-b4f7-64ca8b41e0e9"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "--2022-07-15 14:36:42--  https://s3.amazonaws.com/text-datasets/babi_tasks_1-20_v1-2.tar.gz\n",
            "Resolving s3.amazonaws.com (s3.amazonaws.com)... 54.231.171.72\n",
            "Connecting to s3.amazonaws.com (s3.amazonaws.com)|54.231.171.72|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 11745123 (11M) [application/x-gzip]\n",
            "Saving to: ‘babi_tasks_1-20_v1-2.tar.gz’\n",
            "\n",
            "babi_tasks_1-20_v1- 100%[===================>]  11.20M  6.07MB/s    in 1.8s    \n",
            "\n",
            "2022-07-15 14:36:44 (6.07 MB/s) - ‘babi_tasks_1-20_v1-2.tar.gz’ saved [11745123/11745123]\n",
            "\n",
            "tasks_1-20_v1-2/\n",
            "tasks_1-20_v1-2/en-10k/\n",
            "tasks_1-20_v1-2/en-10k/qa12_conjunction_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa11_basic-coreference_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa4_two-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa20_agents-motivations_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa18_size-reasoning_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa5_three-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa12_conjunction_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa2_two-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa17_positional-reasoning_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa14_time-reasoning_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa6_yes-no-questions_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa17_positional-reasoning_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa15_basic-deduction_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa13_compound-coreference_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa3_three-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa13_compound-coreference_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa15_basic-deduction_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa7_counting_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa11_basic-coreference_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa16_basic-induction_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa20_agents-motivations_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa8_lists-sets_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa19_path-finding_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa4_two-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa7_counting_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa16_basic-induction_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa10_indefinite-knowledge_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa9_simple-negation_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa18_size-reasoning_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa2_two-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa5_three-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa3_three-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa9_simple-negation_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa10_indefinite-knowledge_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa1_single-supporting-fact_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa19_path-finding_test.txt\n",
            "tasks_1-20_v1-2/en-10k/qa14_time-reasoning_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa8_lists-sets_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa6_yes-no-questions_train.txt\n",
            "tasks_1-20_v1-2/en-10k/qa1_single-supporting-fact_train.txt\n",
            "tasks_1-20_v1-2/LICENSE.txt\n",
            "tasks_1-20_v1-2/shuffled/\n",
            "tasks_1-20_v1-2/shuffled/qa12_conjunction_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa11_basic-coreference_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa4_two-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa20_agents-motivations_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa18_size-reasoning_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa5_three-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa12_conjunction_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa2_two-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa17_positional-reasoning_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa14_time-reasoning_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa6_yes-no-questions_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa17_positional-reasoning_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa15_basic-deduction_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa13_compound-coreference_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa3_three-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa13_compound-coreference_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa15_basic-deduction_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa7_counting_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa11_basic-coreference_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa16_basic-induction_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa20_agents-motivations_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa8_lists-sets_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa19_path-finding_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa4_two-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa7_counting_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa16_basic-induction_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa10_indefinite-knowledge_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa9_simple-negation_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa18_size-reasoning_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa2_two-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa5_three-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa3_three-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa9_simple-negation_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa10_indefinite-knowledge_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa1_single-supporting-fact_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa19_path-finding_test.txt\n",
            "tasks_1-20_v1-2/shuffled/qa14_time-reasoning_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa8_lists-sets_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa6_yes-no-questions_train.txt\n",
            "tasks_1-20_v1-2/shuffled/qa1_single-supporting-fact_train.txt\n",
            "tasks_1-20_v1-2/README.txt\n",
            "tasks_1-20_v1-2/hn-10k/\n",
            "tasks_1-20_v1-2/hn-10k/qa12_conjunction_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa11_basic-coreference_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa4_two-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa20_agents-motivations_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa18_size-reasoning_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa5_three-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa12_conjunction_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa2_two-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa17_positional-reasoning_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa14_time-reasoning_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa6_yes-no-questions_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa17_positional-reasoning_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa15_basic-deduction_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa13_compound-coreference_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa3_three-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa13_compound-coreference_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa15_basic-deduction_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa7_counting_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa11_basic-coreference_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa16_basic-induction_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa20_agents-motivations_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa8_lists-sets_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa19_path-finding_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa4_two-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa7_counting_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa16_basic-induction_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa10_indefinite-knowledge_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa9_simple-negation_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa18_size-reasoning_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa2_two-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa5_three-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa3_three-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa9_simple-negation_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa10_indefinite-knowledge_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa1_single-supporting-fact_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa19_path-finding_test.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa14_time-reasoning_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa8_lists-sets_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa6_yes-no-questions_train.txt\n",
            "tasks_1-20_v1-2/hn-10k/qa1_single-supporting-fact_train.txt\n",
            "tasks_1-20_v1-2/hn/\n",
            "tasks_1-20_v1-2/hn/qa12_conjunction_train.txt\n",
            "tasks_1-20_v1-2/hn/qa11_basic-coreference_test.txt\n",
            "tasks_1-20_v1-2/hn/qa4_two-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/hn/qa20_agents-motivations_train.txt\n",
            "tasks_1-20_v1-2/hn/qa18_size-reasoning_test.txt\n",
            "tasks_1-20_v1-2/hn/qa5_three-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/hn/qa12_conjunction_test.txt\n",
            "tasks_1-20_v1-2/hn/qa2_two-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/hn/qa17_positional-reasoning_test.txt\n",
            "tasks_1-20_v1-2/hn/qa14_time-reasoning_test.txt\n",
            "tasks_1-20_v1-2/hn/qa6_yes-no-questions_test.txt\n",
            "tasks_1-20_v1-2/hn/qa17_positional-reasoning_train.txt\n",
            "tasks_1-20_v1-2/hn/qa15_basic-deduction_test.txt\n",
            "tasks_1-20_v1-2/hn/qa13_compound-coreference_train.txt\n",
            "tasks_1-20_v1-2/hn/qa3_three-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/hn/qa13_compound-coreference_test.txt\n",
            "tasks_1-20_v1-2/hn/qa15_basic-deduction_train.txt\n",
            "tasks_1-20_v1-2/hn/qa7_counting_train.txt\n",
            "tasks_1-20_v1-2/hn/qa11_basic-coreference_train.txt\n",
            "tasks_1-20_v1-2/hn/qa16_basic-induction_test.txt\n",
            "tasks_1-20_v1-2/hn/qa20_agents-motivations_test.txt\n",
            "tasks_1-20_v1-2/hn/qa8_lists-sets_test.txt\n",
            "tasks_1-20_v1-2/hn/qa19_path-finding_train.txt\n",
            "tasks_1-20_v1-2/hn/qa4_two-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/hn/qa7_counting_test.txt\n",
            "tasks_1-20_v1-2/hn/qa16_basic-induction_train.txt\n",
            "tasks_1-20_v1-2/hn/qa10_indefinite-knowledge_train.txt\n",
            "tasks_1-20_v1-2/hn/qa9_simple-negation_test.txt\n",
            "tasks_1-20_v1-2/hn/qa18_size-reasoning_train.txt\n",
            "tasks_1-20_v1-2/hn/qa2_two-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/hn/qa5_three-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/hn/qa3_three-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/hn/qa9_simple-negation_train.txt\n",
            "tasks_1-20_v1-2/hn/qa10_indefinite-knowledge_test.txt\n",
            "tasks_1-20_v1-2/hn/qa1_single-supporting-fact_test.txt\n",
            "tasks_1-20_v1-2/hn/qa19_path-finding_test.txt\n",
            "tasks_1-20_v1-2/hn/qa14_time-reasoning_train.txt\n",
            "tasks_1-20_v1-2/hn/qa8_lists-sets_train.txt\n",
            "tasks_1-20_v1-2/hn/qa6_yes-no-questions_train.txt\n",
            "tasks_1-20_v1-2/hn/qa1_single-supporting-fact_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/\n",
            "tasks_1-20_v1-2/shuffled-10k/qa12_conjunction_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa11_basic-coreference_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa4_two-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa20_agents-motivations_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa18_size-reasoning_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa5_three-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa12_conjunction_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa2_two-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa17_positional-reasoning_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa14_time-reasoning_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa6_yes-no-questions_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa17_positional-reasoning_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa15_basic-deduction_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa13_compound-coreference_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa3_three-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa13_compound-coreference_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa15_basic-deduction_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa7_counting_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa11_basic-coreference_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa16_basic-induction_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa20_agents-motivations_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa8_lists-sets_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa19_path-finding_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa4_two-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa7_counting_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa16_basic-induction_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa10_indefinite-knowledge_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa9_simple-negation_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa18_size-reasoning_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa2_two-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa5_three-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa3_three-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa9_simple-negation_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa10_indefinite-knowledge_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa1_single-supporting-fact_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa19_path-finding_test.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa14_time-reasoning_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa8_lists-sets_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa6_yes-no-questions_train.txt\n",
            "tasks_1-20_v1-2/shuffled-10k/qa1_single-supporting-fact_train.txt\n",
            "tasks_1-20_v1-2/en/\n",
            "tasks_1-20_v1-2/en/qa12_conjunction_train.txt\n",
            "tasks_1-20_v1-2/en/qa11_basic-coreference_test.txt\n",
            "tasks_1-20_v1-2/en/qa4_two-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/en/qa20_agents-motivations_train.txt\n",
            "tasks_1-20_v1-2/en/qa18_size-reasoning_test.txt\n",
            "tasks_1-20_v1-2/en/qa5_three-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/en/qa12_conjunction_test.txt\n",
            "tasks_1-20_v1-2/en/qa2_two-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/en/qa17_positional-reasoning_test.txt\n",
            "tasks_1-20_v1-2/en/qa14_time-reasoning_test.txt\n",
            "tasks_1-20_v1-2/en/qa6_yes-no-questions_test.txt\n",
            "tasks_1-20_v1-2/en/qa17_positional-reasoning_train.txt\n",
            "tasks_1-20_v1-2/en/qa15_basic-deduction_test.txt\n",
            "tasks_1-20_v1-2/en/qa13_compound-coreference_train.txt\n",
            "tasks_1-20_v1-2/en/qa3_three-supporting-facts_test.txt\n",
            "tasks_1-20_v1-2/en/qa13_compound-coreference_test.txt\n",
            "tasks_1-20_v1-2/en/qa15_basic-deduction_train.txt\n",
            "tasks_1-20_v1-2/en/qa7_counting_train.txt\n",
            "tasks_1-20_v1-2/en/qa11_basic-coreference_train.txt\n",
            "tasks_1-20_v1-2/en/qa16_basic-induction_test.txt\n",
            "tasks_1-20_v1-2/en/qa20_agents-motivations_test.txt\n",
            "tasks_1-20_v1-2/en/qa8_lists-sets_test.txt\n",
            "tasks_1-20_v1-2/en/qa19_path-finding_train.txt\n",
            "tasks_1-20_v1-2/en/qa4_two-arg-relations_test.txt\n",
            "tasks_1-20_v1-2/en/qa7_counting_test.txt\n",
            "tasks_1-20_v1-2/en/qa16_basic-induction_train.txt\n",
            "tasks_1-20_v1-2/en/qa10_indefinite-knowledge_train.txt\n",
            "tasks_1-20_v1-2/en/qa9_simple-negation_test.txt\n",
            "tasks_1-20_v1-2/en/qa18_size-reasoning_train.txt\n",
            "tasks_1-20_v1-2/en/qa2_two-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/en/qa5_three-arg-relations_train.txt\n",
            "tasks_1-20_v1-2/en/qa3_three-supporting-facts_train.txt\n",
            "tasks_1-20_v1-2/en/qa9_simple-negation_train.txt\n",
            "tasks_1-20_v1-2/en/qa10_indefinite-knowledge_test.txt\n",
            "tasks_1-20_v1-2/en/qa1_single-supporting-fact_test.txt\n",
            "tasks_1-20_v1-2/en/qa19_path-finding_test.txt\n",
            "tasks_1-20_v1-2/en/qa14_time-reasoning_train.txt\n",
            "tasks_1-20_v1-2/en/qa8_lists-sets_train.txt\n",
            "tasks_1-20_v1-2/en/qa6_yes-no-questions_train.txt\n",
            "tasks_1-20_v1-2/en/qa1_single-supporting-fact_train.txt\n"
          ]
        }
      ],
      "source": [
        "!wget https://s3.amazonaws.com/text-datasets/babi_tasks_1-20_v1-2.tar.gz\n",
        "!tar -xzvf babi_tasks_1-20_v1-2.tar.gz"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import keras\n",
        "from keras.preprocessing.text import Tokenizer\n",
        "from keras.preprocessing.sequence import pad_sequences\n",
        "from keras.utils.np_utils import to_categorical\n",
        "import matplotlib.pyplot as plt"
      ],
      "metadata": {
        "id": "LyudwadRPK0c"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "max_vocab=20000\n",
        "emb_dim=15\n",
        "max_sen=2"
      ],
      "metadata": {
        "id": "aeqe5INVQS_8"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_set=open(\"./tasks_1-20_v1-2/en/qa1_single-supporting-fact_train.txt\")\n",
        "test_set=open(\"./tasks_1-20_v1-2/en/qa1_single-supporting-fact_test.txt\")"
      ],
      "metadata": {
        "id": "LyqqWYpsQpUj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def getting_x_y(data):\n",
        "    first_sen_story=[]\n",
        "    second_sen_story=[]\n",
        "    questions=[]\n",
        "    answers=[]\n",
        "    t=1\n",
        "    for line in data:\n",
        "        line=line[2:]\n",
        "        if t==1:\n",
        "            first_sen_story.append(line)\n",
        "            t+=1\n",
        "        elif t==2:\n",
        "            second_sen_story.append(line)\n",
        "            t+=1\n",
        "        else:\n",
        "            q,a,_=line.split(\"\\t\")\n",
        "            questions.append(q)\n",
        "            answers.append(a)\n",
        "            t=1\n",
        "    return first_sen_story,second_sen_story,questions,answers    \n",
        "   "
      ],
      "metadata": {
        "id": "VDGfsRHtQ-04"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_first_sen_story,train_second_sen_story,train_questions,train_answers=getting_x_y(train_set)\n",
        "test_first_sen_story,test_second_sen_story,test_questions,test_answers=getting_x_y(test_set)"
      ],
      "metadata": {
        "id": "jiTqX734RBSx"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "story_tokenizer=Tokenizer(num_words=max_vocab)\n",
        "story_tokenizer.fit_on_texts(train_first_sen_story+train_second_sen_story+test_first_sen_story+test_second_sen_story)\n",
        "train_first_seq_story=story_tokenizer.texts_to_sequences(train_first_sen_story)\n",
        "train_second_seq_story=story_tokenizer.texts_to_sequences(train_second_sen_story)\n",
        "test_first_seq_story=story_tokenizer.texts_to_sequences(test_first_sen_story)\n",
        "test_second_seq_story=story_tokenizer.texts_to_sequences(test_second_sen_story)"
      ],
      "metadata": {
        "id": "X5rfPZLKRD8S"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "max_story_sen_length=max(max(len(s) for s in train_first_seq_story ),max(len(s) for s in train_second_seq_story ),max(len(s) for s in test_first_seq_story ),max(len(s) for s in test_second_seq_story ))"
      ],
      "metadata": {
        "id": "wtfGbs1VRG_2"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_first_seq_story=pad_sequences(train_first_seq_story,maxlen =max_story_sen_length,padding = \"post\")\n",
        "train_second_seq_story=pad_sequences(train_second_seq_story,maxlen =max_story_sen_length,padding = \"post\")\n",
        "test_first_seq_story=pad_sequences(test_first_seq_story,maxlen =max_story_sen_length,padding = \"post\")\n",
        "test_second_seq_story=pad_sequences(test_second_seq_story,maxlen =max_story_sen_length,padding = \"post\")"
      ],
      "metadata": {
        "id": "NjmWqOBiRI5-"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "questions_tokenizer=Tokenizer(num_words=max_vocab)\n",
        "questions_tokenizer.fit_on_texts(train_questions+test_questions)\n",
        "train_seq_questions=questions_tokenizer.texts_to_sequences(train_questions)\n",
        "test_seq_questions=questions_tokenizer.texts_to_sequences(test_questions)"
      ],
      "metadata": {
        "id": "IQfJmaKZRKlX"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "max_questions_sen_length=max(max(len(s) for s in train_seq_questions ),max(len(s) for s in test_seq_questions) )"
      ],
      "metadata": {
        "id": "CXBZz-ZBRMJt"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_seq_questions=pad_sequences(train_seq_questions,maxlen =max_questions_sen_length,padding = \"post\")\n",
        "\n",
        "test_seq_questions=pad_sequences(test_seq_questions,maxlen =max_questions_sen_length,padding = \"post\")\n",
        "\n"
      ],
      "metadata": {
        "id": "fTpMpooUROTH"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "answers_tokenizer=Tokenizer(num_words=max_vocab)\n",
        "answers_tokenizer.fit_on_texts(train_answers+test_answers)\n",
        "train_seq_answers=answers_tokenizer.texts_to_sequences(train_answers)\n",
        "\n",
        "test_seq_answers=answers_tokenizer.texts_to_sequences(test_answers)\n",
        "max_answers_sen_length=max(max(len(s) for s in train_seq_answers ),max(len(s) for s in test_seq_answers) )"
      ],
      "metadata": {
        "id": "J38CSW_5RU_u"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_seq_answers=pad_sequences(train_seq_answers,maxlen =max_answers_sen_length,padding = \"post\")\n",
        "\n",
        "test_seq_answers=pad_sequences(test_seq_answers,maxlen =max_answers_sen_length,padding = \"post\")\n"
      ],
      "metadata": {
        "id": "YFvHAulPRW4F"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def stack_stories(data1,data2):\n",
        "    data=[]\n",
        "    for i in range(len(data1)):\n",
        "        data.append([data1[i],data2[i]])\n",
        "    return np.asarray(data)   "
      ],
      "metadata": {
        "id": "992Vwx_iRZod"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "train_stories=stack_stories(train_first_seq_story,train_second_seq_story)\n",
        "test_stories=stack_stories(test_first_seq_story,test_second_seq_story)"
      ],
      "metadata": {
        "id": "tau48AmVRbrO"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "word2index=answers_tokenizer.word_index\n",
        "train_seq_answers=to_categorical(train_seq_answers)\n",
        "test_seq_answers=to_categorical(test_seq_answers)"
      ],
      "metadata": {
        "id": "L8n3civVRdpl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "story_input=keras.layers.Input(shape=(max_sen,max_story_sen_length,))\n",
        "story_emb=keras.layers.Embedding(max_vocab,emb_dim)\n",
        "x=story_emb(story_input)\n",
        "# the output of x is (2,max_story_sen_length,emb_dim)\n",
        "# model=keras.Model(story_input,x)\n",
        "sum_of_two_stories=keras.layers.Lambda(lambda x:keras.backend.sum(x,axis=2))\n",
        "s=sum_of_two_stories(x)\n",
        "# the output of s is (1,max_story_sen_length,emb_dim)\n",
        "# so its only add them\n",
        "# model1=keras.Model(story_input,s)\n",
        "question_input=keras.layers.Input(shape=(max_questions_sen_length,))\n",
        "question_emb=keras.layers.Embedding(max_vocab,emb_dim)\n",
        "output_of_question_emb=question_emb(question_input)\n",
        "sum_question=keras.layers.Lambda(lambda x:keras.backend.sum(x,axis=1))\n",
        "output_of_sum_question=sum_question(output_of_question_emb)\n",
        "# the output of output_of_question_emb is (1,emb_dim)\n",
        "# so its only add them\n",
        "question_reshape=keras.layers.Reshape((1,emb_dim))\n",
        "output_of_question_reshape=question_reshape(output_of_sum_question)\n",
        "# it reshape them into 3 dim  only\n",
        "# model2=keras.Model(question_input,output_of_sum_question)\n",
        "# model3=keras.Model(question_input,output_of_question_reshape)\n",
        "mul_story_with_question=keras.layers.dot([s,output_of_question_reshape],2)\n",
        "flatten_=keras.layers.Flatten()\n",
        "\n",
        "the_output_flatten=flatten_(mul_story_with_question)\n",
        "# model4=keras.Model([story_input,question_input],mul_story_with_question)\n",
        "# model5=keras.Model([story_input,question_input],the_output_flatten)\n",
        "\n",
        "get_story_weights=keras.layers.Activation('softmax')(the_output_flatten)\n",
        "story_weigths=keras.layers.Reshape((max_sen,1))(get_story_weights)\n",
        "debug_model=keras.Model([story_input,question_input],story_weigths)\n",
        "mul_weight_with_story=keras.layers.dot([story_weigths,s],1)\n",
        "flatten2_=keras.layers.Flatten()\n",
        "y=flatten2_(mul_weight_with_story)\n",
        "y=keras.layers.Dense(len(word2index)+1,activation=\"softmax\")(y)\n",
        "model=keras.Model([story_input,question_input],y)"
      ],
      "metadata": {
        "id": "cvpdwCTORgLU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(\"adam\",loss=keras.losses.CategoricalCrossentropy(),metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "BtE77xn0Rh-C"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "history=model.fit([train_stories,train_seq_questions],train_seq_answers,batch_size=32,epochs=50\n",
        "          ,validation_data=([test_stories,test_seq_questions],test_seq_answers))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MFNlvWXqRkPz",
        "outputId": "4d552f42-9684-41b9-82d3-74a4a3dffdb1"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/50\n",
            "32/32 [==============================] - 2s 17ms/step - loss: 1.9261 - accuracy: 0.2060 - val_loss: 1.8778 - val_accuracy: 0.2980\n",
            "Epoch 2/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.8460 - accuracy: 0.3340 - val_loss: 1.8125 - val_accuracy: 0.3430\n",
            "Epoch 3/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 1.7842 - accuracy: 0.3730 - val_loss: 1.7605 - val_accuracy: 0.3930\n",
            "Epoch 4/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.7306 - accuracy: 0.4260 - val_loss: 1.7144 - val_accuracy: 0.4170\n",
            "Epoch 5/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 1.6806 - accuracy: 0.4440 - val_loss: 1.6718 - val_accuracy: 0.4230\n",
            "Epoch 6/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 1.6336 - accuracy: 0.4480 - val_loss: 1.6321 - val_accuracy: 0.4350\n",
            "Epoch 7/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.5875 - accuracy: 0.4860 - val_loss: 1.5964 - val_accuracy: 0.4530\n",
            "Epoch 8/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.5439 - accuracy: 0.4920 - val_loss: 1.5580 - val_accuracy: 0.4600\n",
            "Epoch 9/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.4983 - accuracy: 0.5120 - val_loss: 1.5215 - val_accuracy: 0.4780\n",
            "Epoch 10/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.4525 - accuracy: 0.5210 - val_loss: 1.4846 - val_accuracy: 0.4970\n",
            "Epoch 11/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 1.4075 - accuracy: 0.5590 - val_loss: 1.4431 - val_accuracy: 0.5220\n",
            "Epoch 12/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 1.3576 - accuracy: 0.5800 - val_loss: 1.4034 - val_accuracy: 0.5500\n",
            "Epoch 13/50\n",
            "32/32 [==============================] - 0s 11ms/step - loss: 1.3051 - accuracy: 0.6110 - val_loss: 1.3593 - val_accuracy: 0.5730\n",
            "Epoch 14/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 1.2532 - accuracy: 0.6200 - val_loss: 1.3126 - val_accuracy: 0.5850\n",
            "Epoch 15/50\n",
            "32/32 [==============================] - 0s 11ms/step - loss: 1.1960 - accuracy: 0.6470 - val_loss: 1.2710 - val_accuracy: 0.5970\n",
            "Epoch 16/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.1419 - accuracy: 0.6750 - val_loss: 1.2273 - val_accuracy: 0.6100\n",
            "Epoch 17/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.0908 - accuracy: 0.6870 - val_loss: 1.1875 - val_accuracy: 0.6200\n",
            "Epoch 18/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.0449 - accuracy: 0.7000 - val_loss: 1.1509 - val_accuracy: 0.6490\n",
            "Epoch 19/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 1.0069 - accuracy: 0.7050 - val_loss: 1.1240 - val_accuracy: 0.6600\n",
            "Epoch 20/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.9732 - accuracy: 0.7100 - val_loss: 1.0985 - val_accuracy: 0.6660\n",
            "Epoch 21/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.9444 - accuracy: 0.7150 - val_loss: 1.0790 - val_accuracy: 0.6750\n",
            "Epoch 22/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.9213 - accuracy: 0.7170 - val_loss: 1.0620 - val_accuracy: 0.6760\n",
            "Epoch 23/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.9028 - accuracy: 0.7210 - val_loss: 1.0482 - val_accuracy: 0.6770\n",
            "Epoch 24/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.8877 - accuracy: 0.7240 - val_loss: 1.0407 - val_accuracy: 0.6740\n",
            "Epoch 25/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.8759 - accuracy: 0.7240 - val_loss: 1.0341 - val_accuracy: 0.6740\n",
            "Epoch 26/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8655 - accuracy: 0.7260 - val_loss: 1.0271 - val_accuracy: 0.6770\n",
            "Epoch 27/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8574 - accuracy: 0.7230 - val_loss: 1.0242 - val_accuracy: 0.6740\n",
            "Epoch 28/50\n",
            "32/32 [==============================] - 0s 11ms/step - loss: 0.8486 - accuracy: 0.7290 - val_loss: 1.0172 - val_accuracy: 0.6750\n",
            "Epoch 29/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.8423 - accuracy: 0.7320 - val_loss: 1.0138 - val_accuracy: 0.6770\n",
            "Epoch 30/50\n",
            "32/32 [==============================] - 0s 11ms/step - loss: 0.8357 - accuracy: 0.7370 - val_loss: 1.0121 - val_accuracy: 0.6770\n",
            "Epoch 31/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8312 - accuracy: 0.7350 - val_loss: 1.0086 - val_accuracy: 0.6770\n",
            "Epoch 32/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8277 - accuracy: 0.7370 - val_loss: 1.0071 - val_accuracy: 0.6730\n",
            "Epoch 33/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8211 - accuracy: 0.7380 - val_loss: 1.0020 - val_accuracy: 0.6700\n",
            "Epoch 34/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8170 - accuracy: 0.7420 - val_loss: 1.0007 - val_accuracy: 0.6680\n",
            "Epoch 35/50\n",
            "32/32 [==============================] - 0s 11ms/step - loss: 0.8141 - accuracy: 0.7360 - val_loss: 0.9984 - val_accuracy: 0.6710\n",
            "Epoch 36/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8104 - accuracy: 0.7430 - val_loss: 0.9968 - val_accuracy: 0.6710\n",
            "Epoch 37/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8068 - accuracy: 0.7450 - val_loss: 0.9943 - val_accuracy: 0.6700\n",
            "Epoch 38/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.8038 - accuracy: 0.7440 - val_loss: 0.9928 - val_accuracy: 0.6700\n",
            "Epoch 39/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.8000 - accuracy: 0.7470 - val_loss: 0.9919 - val_accuracy: 0.6700\n",
            "Epoch 40/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.7987 - accuracy: 0.7460 - val_loss: 0.9885 - val_accuracy: 0.6710\n",
            "Epoch 41/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.7956 - accuracy: 0.7460 - val_loss: 0.9875 - val_accuracy: 0.6640\n",
            "Epoch 42/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.7942 - accuracy: 0.7470 - val_loss: 0.9871 - val_accuracy: 0.6660\n",
            "Epoch 43/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.7913 - accuracy: 0.7460 - val_loss: 0.9853 - val_accuracy: 0.6690\n",
            "Epoch 44/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.7898 - accuracy: 0.7490 - val_loss: 0.9825 - val_accuracy: 0.6690\n",
            "Epoch 45/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.7902 - accuracy: 0.7460 - val_loss: 0.9821 - val_accuracy: 0.6700\n",
            "Epoch 46/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.7864 - accuracy: 0.7480 - val_loss: 0.9821 - val_accuracy: 0.6700\n",
            "Epoch 47/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.7851 - accuracy: 0.7480 - val_loss: 0.9797 - val_accuracy: 0.6700\n",
            "Epoch 48/50\n",
            "32/32 [==============================] - 0s 11ms/step - loss: 0.7832 - accuracy: 0.7430 - val_loss: 0.9793 - val_accuracy: 0.6680\n",
            "Epoch 49/50\n",
            "32/32 [==============================] - 0s 10ms/step - loss: 0.7829 - accuracy: 0.7510 - val_loss: 0.9784 - val_accuracy: 0.6700\n",
            "Epoch 50/50\n",
            "32/32 [==============================] - 0s 9ms/step - loss: 0.7815 - accuracy: 0.7460 - val_loss: 0.9786 - val_accuracy: 0.6690\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history[\"accuracy\"],label=\"accuracy\")\n",
        "plt.plot(history.history[\"val_accuracy\"],label=\"val_accuracy\")\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "KBD-RlU7RnZV",
        "outputId": "89d3efa4-0384-44e0-c126-fa52a22329c8"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(history.history[\"loss\"],label=\"loss\")\n",
        "plt.plot(history.history[\"val_loss\"],label=\"loss\")\n",
        "plt.legend()\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "yymJQFlhRumM",
        "outputId": "38e86ce0-bcae-4369-b9e9-cefd6eef76bb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "i=np.random.choice(len(train_first_sen_story))\n",
        "z=model.predict([train_stories[i:i+1],train_seq_questions[i:i+1]])\n",
        "idx=np.argmax( z[0,:])\n",
        "print(train_first_sen_story[i])\n",
        "print(train_second_sen_story[i])\n",
        "print(train_questions[i])\n",
        "answers_tokenizer.index_word[idx]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 125
        },
        "id": "vIQRIiaBRx7i",
        "outputId": "89c81b44-0ea6-40cd-89ba-63515f8cd347"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Daniel moved to the bathroom.\n",
            "\n",
            "Mary journeyed to the garden.\n",
            "\n",
            "Where is Daniel? \n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'bathroom'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "0p_p_nJCR1OX"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}