MoviCHTRv3.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<img src=\"MoviCHTR logo.png\" width=\"500\" height=\"500\">"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# MoviCHTR\n",
    "\n",
    "### A conversational chat bot for all (SOME) things movies"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 1. Project Setup\n",
    "### The first step will be to create an initial project we can build up\n",
    "### For more information for the path I chose please check documentation: https://rasa.com/docs/rasa/api/jupyter-notebooks/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#loading libraries\n",
    "#Since Rasa runs asynchronous code and Jupyter is already run on event loops,\n",
    "#we need to add nest_asyncio (allows event loops to be nested)\n",
    "import nest_asyncio\n",
    "nest_asyncio.apply()\n",
    "\n",
    "#For now we will ignore many of the warnings\n",
    "import logging, io, json, warnings\n",
    "logging.basicConfig(level=\"CRITICAL\")\n",
    "\n",
    "#these are the modules needed to create a simple project and the initial folders\n",
    "import rasa\n",
    "from rasa.cli.scaffold import create_initial_project\n",
    "import os\n",
    "from rasa.jupyter import chat\n",
    "\n",
    "#for the puposes of training just the nlu we will need metadata for the json files and and interpreter for parsing\n",
    "from rasa.nlu.model import Metadata, Interpreter\n",
    "import tarfile\n",
    "\n",
    "# modules for nlu training\n",
    "from rasa.nlu.training_data import load_data\n",
    "from rasa.nlu.config import RasaNLUModelConfig\n",
    "from rasa.nlu.model import Trainer\n",
    "from rasa.nlu import config\n",
    "\n",
    "#training rasa core\n",
    "from rasa.core.policies.keras_policy import KerasPolicy\n",
    "from rasa.core.policies.memoization import MemoizationPolicy\n",
    "from rasa.core.policies.mapping_policy import MappingPolicy\n",
    "from rasa.core.agent import Agent\n",
    "import asyncio\n",
    "\n",
    "#deploying bot and loading assistant\n",
    "import IPython\n",
    "from IPython.display import clear_output\n",
    "from rasa.core.agent import Agent\n",
    "from rasa.core.interpreter import NaturalLanguageInterpreter\n",
    "from rasa.utils.endpoints import EndpointConfig\n",
    "from IPython.display import Image\n",
    "from rasa.core.channels.slack import SlackInput\n",
    "import time\n",
    "\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Defining file paths"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "#init the variable for our project\n",
    "movichtr_project = \"MoviCHTR\"\n",
    "movichtr_config_path = movichtr_project + \"/config.yml\"\n",
    "movichtr_training_files = movichtr_project + \"/data/\"\n",
    "movichtr_training_file_nlu = movichtr_project + \"/data/nlu.md\"\n",
    "movichtr_training_file_stories = movichtr_project + \"/data/stories.md\"\n",
    "movichtr_domain_path = movichtr_project + \"/domain.yml\"\n",
    "movichtr_models_output = movichtr_project + \"/models/\"\n",
    "movichtr_actions_file = movichtr_project + \"/actions.py\"\n",
    "movichtr_endpoints_file = movichtr_project + \"/endpoints.yml\"\n",
    "movichtr_model_name = \"movichtr_rasa_model\"\n",
    "movichtr_credentials_path = movichtr_project + \"/credentials.yml\"\n",
    "movichtr_api_file = movichtr_project + \"/api_handler.py\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Setting up a global variable for the rasa server port"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "#in case we are not using default ports here is a an effective global variable to change ports on the go\n",
    "PORT = 5055"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Method for project init\n",
    "\n",
    "#### Only needs to be run the first time "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#### change this cell into code for the first project setup, no need to be repeated in further runs\n",
    "create_initial_project(movichtr_project)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Core stories/configuration did not change. No need to retrain Core model.\n",
      "Training NLU model...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epochs: 100%|█████████████████████████████████████████████████| 300/300 [00:35<00:00,  8.35it/s, loss=0.494, acc=0.995]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "NLU model training completed.\n",
      "Your Rasa model is trained and saved at 'C:\\Users\\sousa\\Desktop\\MoviCHTRv3\\MoviCHTR\\models\\movichtr_rasa_model.tar.gz'.\n"
     ]
    }
   ],
   "source": [
    "#### training with the populated file tree not essential but should be run once we create the first project for a check\n",
    "movichtr_model_path = rasa.train(movichtr_domain_path, movichtr_config_path, [movichtr_training_files], movichtr_models_output, fixed_model_name=movichtr_model_name)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "#defining a function to allow writing to file \n",
    "def write_to_file(file, string_to_write):\n",
    "    file = open(file, \"w\")\n",
    "    file.write(string_to_write)\n",
    "    file.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Customizing the config file"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### This is a now dormant config file\n",
    "### In my experience using anything based in Spacy with less than 300 examples for the training data causes issues with the slot filling\n",
    "### I leave it available for your testing\n",
    "### surprisingly the default supervised embedings worked better for me\n",
    "\n",
    "movichtr_config = \"\"\"\n",
    "language: en\n",
    "\n",
    "pipeline:\n",
    "- name: SpacyNLP\n",
    "- name: SpacyTokenizer\n",
    "- name: SpacyFeaturizer\n",
    "- name: RegexFeaturizer\n",
    "- name: CRFEntityExtractor\n",
    "- name: EntitySynonymMapper\n",
    "- name: SklearnIntentClassifier\n",
    "- name: CountVectorsFeaturizer\n",
    "  lowercase: true\n",
    "  token_pattern: (?u)\\b\\w+\\b\n",
    "  \n",
    "policies:\n",
    "- name: MemoizationPolicy\n",
    "- name: KerasPolicy\n",
    "- name: MappingPolicy\n",
    "\n",
    "\"\"\"\n",
    "write_to_file(movichtr_config_path, movichtr_config)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### This is a tensorflow based training config\n",
    "movichtr_config = \"\"\"\n",
    "language: en\n",
    "pipeline: \n",
    "- name: \"WhitespaceTokenizer\"\n",
    "  intent_split_symbol: \" \"\n",
    "- name: \"CountVectorsFeaturizer\"\n",
    "- name: \"EmbeddingIntentClassifier\"\n",
    "\n",
    "policies:\n",
    "- name: MemoizationPolicy\n",
    "- name: KerasPolicy\n",
    "- name: MappingPolicy\n",
    "\n",
    "\"\"\"\n",
    "write_to_file(movichtr_config_path, movichtr_config)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### This is the default config with supervised embedings\n",
    "movichtr_config = \"\"\"\n",
    "language: en\n",
    "pipeline: supervised_embeddings\n",
    "\n",
    "policies:\n",
    "- name: MemoizationPolicy\n",
    "- name: KerasPolicy\n",
    "- name: MappingPolicy\n",
    "\n",
    "\"\"\"\n",
    "write_to_file(movichtr_config_path, movichtr_config)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 2. Rasa Natural Understanding\n",
    "## Rasa NLU extracts entities and intents\n",
    "\n",
    "### Rasa NLU is an open-source natural language processing tool for intent classification, response retrieval and entity extraction in chatbots\n",
    "### the train function run above generates a gunzip file with the models\n",
    "### On the next steps we will extract the files and generate an interpreter that can be used to visualize the output of Rasa NLU"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'intent': {'name': 'greet', 'confidence': 0.9824914336204529},\n",
       " 'entities': [],\n",
       " 'intent_ranking': [{'name': 'greet', 'confidence': 0.9824914336204529},\n",
       "  {'name': 'bye', 'confidence': 0.006811661645770073},\n",
       "  {'name': 'deny', 'confidence': 0.003710784250870347},\n",
       "  {'name': 'search_movie_director', 'confidence': 0.0014460603706538677},\n",
       "  {'name': 'goodbye', 'confidence': 0.0012023670133203268},\n",
       "  {'name': 'movie_search_by_director', 'confidence': 0.0011557255638763309},\n",
       "  {'name': 'search_rating_for_movie', 'confidence': 0.000881293963175267},\n",
       "  {'name': 'search_popularity_by_year', 'confidence': 0.0006107864319346845},\n",
       "  {'name': 'affirm', 'confidence': 0.0005996540421620011},\n",
       "  {'name': 'get_introduction', 'confidence': 0.00038566571311093867}],\n",
       " 'text': 'hi'}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# opening the initiallly created model files and testing the interpreter\n",
    "\n",
    "tar = tarfile.open(movichtr_models_output + movichtr_model_name + \".tar.gz\", \"r:gz\")\n",
    "tar.extractall(movichtr_models_output + movichtr_model_name)\n",
    "tar.close()\n",
    "\n",
    "interpreter = Interpreter.load(movichtr_models_output + movichtr_model_name + \"/nlu/\")\n",
    "\n",
    "interpreter.parse(\"hi\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### The output above shows us what information the system extracts from the sentence passed:\n",
    "\n",
    "- intent: corresponds to the intent the system believes is being conveyed in the message;\n",
    "- entities: a list that contains relevant entities in the sentence (empty in this case);\n",
    "- intent_ranking: list of all intents ordered by confidence.  Each intent has the following format:\n",
    "   -First line in the format ## intent:{name_of_intent};\n",
    "   -N lines, each corresponding to a sentence that conveys the intent preceded by -;\n",
    "   -Blank line.\n",
    "   \n",
    "### The code to visualized the training file without navigating to the folder is as follows:\n",
    "f = open(movichtr_training_file_nlu)\n",
    "content = f.read()\n",
    "print(content)\n",
    "f.close()  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### For the next few steps I will create the elements that will help prepare data for information extration using nlu\n",
    "\n",
    "### For a brief overview on these elements please check this small article: https://towardsdatascience.com/a-beginners-guide-to-rasa-nlu-for-intent-classification-and-named-entity-recognition-a4f0f76b2a96"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## nlu.md data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "movichtr_intents = \"\"\"\n",
    "## intent:greet\n",
    "- hey\n",
    "- hello\n",
    "- hi\n",
    "- good morning\n",
    "- good evening\n",
    "- hey there\n",
    "\n",
    "## intent:goodbye\n",
    "- bye\n",
    "- goodbye\n",
    "- see you around\n",
    "- see you later\n",
    "\n",
    "## intent:affirm\n",
    "- yes\n",
    "- indeed\n",
    "- of course\n",
    "- that sounds good\n",
    "- correct\n",
    "\n",
    "## intent:deny\n",
    "- no\n",
    "- never\n",
    "- I don't think so\n",
    "- don't like that\n",
    "- no way\n",
    "- not really\n",
    "\n",
    "## intent:bye\n",
    "- No, I am good as of now. Bye\n",
    "- Bye\n",
    "- Bbye\n",
    "\n",
    "## intent:get_introduction\n",
    "- Tell me more about you.\n",
    "- What do you do?\n",
    "- how to use you?\n",
    "- instructions\n",
    "- How do you work?\n",
    "\n",
    "## intent:movie_search_by_actor\n",
    "- Find movie with [John Travolta](actor) and [Uma Thurman](actor)\n",
    "- Find movie with [Brad Pitt](actor)\n",
    "- Find movie with [Johny Depp](actor)\n",
    "- In which movie [Johny Depp](actor) shot?\n",
    "- What is the movie with [Brad Pitt](actor)\n",
    "- Hey bot find movies with [Woody Allen](actor)\n",
    "- Find me a movie with [Madonna](actor)\n",
    "- Give me a movie with [Sting](actor)\n",
    "- where did [Anna Nab](actor) star in?\n",
    "- Movies with [Daniel Day-Lewis](actor)\n",
    "- Where did [Robert De Niro](actor) star in?\n",
    "- where did [Philip Seymour Hoffman](actor) act in\n",
    "- find movies with [Meryl Streep](actor)\n",
    "- give me movies with [Adam Driver](actor)\n",
    "- find me movies with [Jennifer Lawrence](actor)\n",
    "- well, find me movies with [Jennifer Lawrence](actor)\n",
    "- Find me movies with [Sting](actor)\n",
    "- Find me movies with [Jennifer Lawrence](actor)\n",
    "- movies performed by [jean reno](actor)\n",
    "- Find me movies with [Brad Pitt](actor)\n",
    "- movies with [Jean Reno](actor)\n",
    "- give me movies by [Brad Pitt](actor)\n",
    "\n",
    "## intent:movie_search_by_director\n",
    "- Find a movie directed by [Quentin Tarantino](director).\n",
    "- What can you find me directed by [Guillermo del Toro](director?\n",
    "- What has [Ridley Scott](director) directed?\n",
    "- Find titles by [Oliver Stone](director).\n",
    "- Give me movies by [Martin Scorcese](director).\n",
    "- Find me movies by [Ingmar Bergman](director)\n",
    "- what does [justin benson](director) have out there?\n",
    "- give me titles by [Aaron Moorhead](director)\n",
    "- titles by [some dude](director)\n",
    "- movie by [beach dude from ocean beach](director)\n",
    "- movie directed by [Francis Ford Coppola](director)\n",
    "- Find movies by [Oliver Stone](director)\n",
    "- movies directed by [Woody Allen](director)\n",
    "- movies by [Oliver Stone](director)\n",
    "- movies by [Woody Allen](director)\n",
    "- What is the director [Woody Allen](director) known for\n",
    "- find movies by [Woody Allen](director)\n",
    "- Get me a movie by [Oliver Stone](director)\n",
    "- get me a movie by [Woody Allen](director)\n",
    "- give me movies by [Stanley Kubrick](director)\n",
    "- Find me movies by [Stanley Kubrick](director)\n",
    "- Find me movies by [Stanley Kubrik](director)\n",
    "- give me movies by [Guillermo del Toro](director)\n",
    "- [guy ritchie](director) director\n",
    "- What has [Ridley Scott](director) directed\n",
    "- Give me a movie by [Guillermo del Toro](director)\n",
    "- Find me a movie by [Oliver Stone](director)\n",
    "- find me movies by [Oliver Stone](director)\n",
    "- What is the director [Woody Allen](director) known for\n",
    "\n",
    "## intent:release_year_search_by_movie\n",
    "- When was [Alien](movie_title) released?\n",
    "- What year did [Pulp Fiction](movie_title) come out?\n",
    "- When was [Zombieland](movie_title) out?\n",
    "- what is the release of [Terminator: Dark Fate](movie_title)\n",
    "- when was [The Dead Do not Die](movie_title) out?\n",
    "- give me the release of [Tigers Are Not Afraid](movie_title)\n",
    "- When was [Avengers: Endgame](movie_title) released?\n",
    "- What was [The devil wears Prada](movie_title) released\n",
    "- out date for [black panther](movie_title)\n",
    "- release date for [Roma](movie_title)\n",
    "- when was [The Mule](movie_title) released?\n",
    "- let me know when was [I am not a witch](movie_title) released?\n",
    "- give me the release for [the favourite](movie_title).\n",
    "- release for [a house in the prairie](movie_title)\n",
    "- when was [Match point](movie_title) released\n",
    "- when was [The Lure](movie_title) released\n",
    "- when was [Joker](movie_title) released?\n",
    "- when was [Joker](movie_title) released\n",
    "- when was [joker](movie_title) released\n",
    "- when was [Alien](movie_title) released?\n",
    "- When was [The Matrix](movie_title) released?\n",
    "- when was [The Matrix](movie_title) released\n",
    "- When was [The Matrix Reloaded](movie_title) released\n",
    "- when was [The devil wears prada](movie_title) released\n",
    "- when was [alien](movie_title) released\n",
    "- When was [Excess Baggage](movie_title) released\n",
    "- when was [The Shining](movie_title) released\n",
    "- when was [Alien](movie_title) released\n",
    "- When was [se7en](movie_title) released\n",
    "- when was [the shining](movie_title) released\n",
    "- When was [Fight Club](movie_title) released\n",
    "- When was [Alien](movie_title) released\n",
    "- What was the release of [Patriot Games](movie_title)\n",
    "- when was [The Shining](movie_title) released\n",
    "- When was [Patriot Games](movie_title) released\n",
    "- When was [Terminator Salvation](movie_title) released\n",
    "\n",
    "## intent:search_overview_for_movie\n",
    "- What is the plot for [Alien](movie_title)?\n",
    "- give me the summary for [Terminator Genesys](movie_title)\n",
    "- overview for [Terminator: Dark Fate](movie_title)\n",
    "- give me the overview for [The Art of Self Defense](movie_title)\n",
    "- what is [The Dead Do not Die](movie_title) about?\n",
    "- give me the plot for [Match Point](movie_title)\n",
    "- What is [Kramer vs Kramer](movie_title) about?\n",
    "- give me the summary of [Maleficient: Mistress of Evil](movie_title).\n",
    "- plot for [skills matter](movie_title)\n",
    "- overview for [hp laptop](movie_title)\n",
    "- what is the plot for [Coco](movie_title)\n",
    "\n",
    "## intent:search_rating_for_movie\n",
    "- What is the rating for [Alien](movie_title)\n",
    "- How was [Maleficient: Mistress of Evil](movie_title) rated\n",
    "- How did [Terminator: Dark Fate](movie_title) rate\n",
    "- what was [Avengers: Endgame](movie_title) rated with?\n",
    "- give me the rating for [Spider-Man: Far from home](movie_title)\n",
    "- rating for [The Lure](movie_title)\n",
    "- give me the rating for [Thelma](movie_title)\n",
    "- score for [city of ghosts](movie_title)\n",
    "- how is [super dark times](movie_title) rated?\n",
    "\n",
    "## intent:search_popularity_by_year\n",
    "- What were the popular movies for [2016](release)\n",
    "- Which movies were more popular in [1976](release)\n",
    "- give me the most popular movies of [1983](release)\n",
    "- what was popular in [2000](release)\n",
    "- Give popular movies of [1999](release)\n",
    "\n",
    "## intent:search_movie_director\n",
    "- Who directed [Black Panther](movie_title)\n",
    "- who was the director of [I am not a witch](movie_title)\n",
    "- Get me the director of [Mission Impossible Fallout](movie_title)\n",
    "- [Leave No Trace](movie_title) was by who?\n",
    "- Who directed [Terminator](movie_title)\n",
    "- [Joker](movie_title) was directed by who?\n",
    "- who was [Paddington 2](movie_title) by?\n",
    "- [Filmworker](movie_title) director.\n",
    "- who directed [The Ballad of Buster Scruggs](movie_title)?\n",
    "- who directed [pulp fiction](movie_title)\n",
    "- Who directed [Match Point](movie_title)\n",
    "- who directed [Coco](movie_title)\n",
    "- Let me know the director of [Matrix](movie_title)\n",
    "- who directed [New York New York](movie_title)?\n",
    "- Who directed the film [Snatch](movie_title)?\n",
    "\n",
    "## intent:search_movie_actors\n",
    "- Who starred in [It Follows](movie_title)?\n",
    "- Which actors were in [Ex Machina](movie_title)\n",
    "- stars in [The revenant](movie_title)?\n",
    "- Who acted in [What We Do in the Shadows](movie_title)\n",
    "- actors in [Sicario](movie_title)\n",
    "- acting in [The big short](movie_title)\n",
    "- Starring in [Spotlight](movie_title)\n",
    "- who starred in [The Hateful Eight](movie_title)\n",
    "- which actors in [the martian](movie_title)\n",
    "- Stars in [Slow West](movie_title)\n",
    "- who acted in [Inside Out](movie_title)\n",
    "- who acted in [Inside Out](movie_title)\n",
    "- who acted in [Pulp Fiction](movie_title)\n",
    "- Who acted in [Match Point](movie_title)\n",
    "- who acted in [Alien](movie_title)\n",
    "- Who acted in [Alien](movie_title)\n",
    "- Who acted in [Alien](movie_title)?\n",
    "- who acted in [alien](movie_title)\n",
    "- who acted in [sleepy hollow](movie_title)\n",
    "- who acted in [coco](movie_title)?\n",
    "- who acted in [bumblebee](movie_title)\n",
    "- who acted in [coco](movie_title)\n",
    "- who acted in [Frozen](movie_title)\n",
    "- who acted in [Fight Club](movie_title)\n",
    "- actors of [deep blue](movie_title)\n",
    "\n",
    "## intent:search_similar_movies\n",
    "- movies similar to [Heat](movie_title)\n",
    "- which movies are similar to [The Silence of the Lambs](movie_title)\n",
    "- suggest something like [No Country for Old Men](movie_title)\n",
    "- give me something like [Gangs of New York](movie_title)\n",
    "- What is similar to [The Lord of the Rings The Two Towers](movie_title)\n",
    "- movie like [Mystic River](movie_title)\n",
    "- suggestion like [Ace Ventura Pet Detective](movie_title)\n",
    "- similar movies to [coco](movie_title)\n",
    "- similar movies to [Alien](movie_title)\n",
    "- suggest movies like [Black Panther](movie_title)\n",
    "- Can you suggest movies similar to [deep blue](movie_title)\n",
    "\n",
    "##synonym:movie_title\n",
    "-flick\n",
    "-film\n",
    "-picture\n",
    "-motion picture \n",
    "-feature film\n",
    "-silver screen\n",
    "-titles\n",
    "\n",
    "## regex:release\n",
    "- [0-9]{4}\n",
    "\n",
    "## lookup:movie_title\n",
    "- Alien\n",
    "- Zombieland\n",
    "- Coco\n",
    "- The Silence of the Lambs\n",
    "- Ace Ventura Pet Detective\n",
    "- Mystic River\n",
    "- Gangs of New York\n",
    "- The Lord of the Rings The Two Towers\n",
    "- Heat\n",
    "- No Country for Old Men\n",
    "- Match Point\n",
    "- The Hateful Eight\n",
    "- Sicario\n",
    "- Inside Out\n",
    "- Slow West\n",
    "- The revenant\n",
    "- Ex Machina\n",
    "- The big short\n",
    "- The Ballad of Buster Scruggs\n",
    "- Paddington 2\n",
    "- Filmworker\n",
    "- Leave No Trace\n",
    "- Joker\n",
    "- Mission Impossible Fallout\n",
    "- I am not a witch\n",
    "- Maleficient Mistress of Evil\n",
    "- Avengers Endgame\n",
    "- Se7en\n",
    "- Blade Runner 2049\n",
    "- Spider-Man: Far from home\n",
    "- Tigers Are Not Afraid\n",
    "\n",
    "## lookup:actors\n",
    "- John Travolta\n",
    "- Brad Pitt\n",
    "- Johny Depp\n",
    "- Woody Allen\n",
    "- Madonna\n",
    "- Sting\n",
    "- Anna Nab\n",
    "- Daniel Day-Lewis\n",
    "- Robert De Niro\n",
    "- Philip Seymour Hoffman\n",
    "\n",
    "## lookup:director\n",
    "- Quentin Tarantino\n",
    "- Guillermo del Toro\n",
    "- Ridley Scott\n",
    "- Oliver Stone\n",
    "- Martin Scorcese\n",
    "- Ingmar Bergman\n",
    "- justin benson\n",
    "- Aaron Moorhead\n",
    "- some dude\n",
    "- beach dude from ocean beach\n",
    "- Francis Ford Coppola\n",
    "\n",
    "\"\"\"\n",
    "\n",
    "write_to_file(movichtr_training_file_nlu, movichtr_intents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_movichtr_nlu_model():\n",
    "    movichtr_training_data = load_data(movichtr_training_file_nlu)\n",
    "\n",
    "    trainer = Trainer(config.load(movichtr_config_path))\n",
    "\n",
    "    interpreter = trainer.train(movichtr_training_data)\n",
    "    \n",
    "    model_directory = trainer.persist(movichtr_models_output + movichtr_model_name, fixed_model_name=\"nlu\")\n",
    "    \n",
    "    return interpreter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Epochs: 100%|█████████████████████████████████████████████████| 300/300 [00:33<00:00,  9.09it/s, loss=0.511, acc=1.000]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "{'intent': {'name': 'movie_search_by_actor', 'confidence': 0.9117981791496277},\n",
       " 'entities': [{'start': 20,\n",
       "   'end': 33,\n",
       "   'value': 'Frank Sinatra',\n",
       "   'entity': 'actor',\n",
       "   'confidence': 0.5626121699294669,\n",
       "   'extractor': 'CRFEntityExtractor'}],\n",
       " 'intent_ranking': [{'name': 'movie_search_by_actor',\n",
       "   'confidence': 0.9117981791496277},\n",
       "  {'name': 'movie_search_by_director', 'confidence': 0.026723694056272507},\n",
       "  {'name': 'get_introduction', 'confidence': 0.01876167207956314},\n",
       "  {'name': 'search_similar_movies', 'confidence': 0.009204014204442501},\n",
       "  {'name': 'search_overview_for_movie', 'confidence': 0.008875792846083641},\n",
       "  {'name': 'search_rating_for_movie', 'confidence': 0.0060564251616597176},\n",
       "  {'name': 'search_movie_actors', 'confidence': 0.004130893386900425},\n",
       "  {'name': 'goodbye', 'confidence': 0.0036785416305065155},\n",
       "  {'name': 'search_popularity_by_year', 'confidence': 0.0034548151306807995},\n",
       "  {'name': 'release_year_search_by_movie',\n",
       "   'confidence': 0.002727022860199213}],\n",
       " 'text': \"What's a movie with Frank Sinatra?\"}"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#testing the interpreter\n",
    "interpreter = train_movichtr_nlu_model()\n",
    "display(interpreter.parse(\"What's a movie with Frank Sinatra?\")) "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# 3. Rasa Core\n",
    "\n",
    "##  Rasa Core handles the conversation flow, utterances and actions\n",
    "\n",
    "Now that we added some capabilities to the NLU system that our bot is going to use, we have to adjust and improve the way it can dialog with a user.\n",
    "\n",
    "Since we have not changed the dialog and trained the rasa core yet, the bot remains roughly the same as before. To build upon this we will add a few elements:\n",
    "\n",
    " - stories: A story is a representation of a conversation between a user and an AI assistant, converted into a specific format where user inputs are expressed as corresponding intents (and entities where necessary) while the responses of an assistant are expressed as corresponding action names. A training example for the Rasa Core dialogue system is called a story.To read more about stories: https://rasa.com/docs/rasa/core/stories/\n",
    " \n",
    " - domains: The Domain defines the universe in which your assistant operates. It specifies the intents, entities, slots, and actions your bot should know about. Optionally, it can also include templates for the things your bot can say. Follow the link for more exhaustive information: https://rasa.com/docs/rasa/core/domains/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "movichtr_model_path = movichtr_models_output + movichtr_model_name + \"/\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "movichtr_stories = \"\"\"\n",
    "## fallback\n",
    "  - utter_greet\n",
    "\n",
    "## greeting path 1\n",
    "* greet\n",
    "  - utter_greet\n",
    "\n",
    "## get movie by actor\n",
    "* movie_search_by_actor\n",
    "  - utter_movie_search_by_actor\n",
    "  - action_movie_search_by_actor\n",
    "\n",
    "## get movie by director\n",
    "* movie_search_by_director\n",
    "  - utter_movie_search_by_director\n",
    "  - action_movie_search_by_director\n",
    "\n",
    "## get release year for given movie\n",
    "* release_year_search_by_movie\n",
    "  - utter_search_release_year_for_movie\n",
    "  - action_search_release_year_for_movie\n",
    "\n",
    "## get overview for movie\n",
    "* search_overview_for_movie\n",
    "  - utter_search_overview_for_movie\n",
    "  - action_search_overview_for_movie\n",
    "  \n",
    "## get rating for movie\n",
    "* search_rating_for_movie\n",
    "  - utter_search_rating_for_movie\n",
    "  - action_search_rating_for_movie\n",
    "\n",
    "## get popular movies by year\n",
    "* search_popularity_by_year\n",
    "  - utter_search_popularity_by_year\n",
    "  - action_search_popularity_by_year\n",
    "\n",
    "## get movie director \n",
    "* search_movie_director\n",
    "  - utter_search_movie_director\n",
    "  - action_search_movie_director\n",
    "  \n",
    "## get movie actors\n",
    "* search_movie_actors\n",
    "  - utter_search_movie_actors\n",
    "  - action_search_movie_actors\n",
    "  \n",
    "## get similar movies\n",
    "* search_similar_movies\n",
    "  - utter_search_similar_movies\n",
    "  - action_search_similar_movies\n",
    "\n",
    "## get introduction\n",
    "* get_introduction\n",
    "  - utter_introduction\n",
    "\n",
    "## bye path 1\n",
    "* bye\n",
    "  - utter_bye\n",
    "  \n",
    "## happy path 1\n",
    "* greet\n",
    "  - utter_greet\n",
    "* get_introduction\n",
    "  - utter_introduction\n",
    "* movie_search_by_director{\"director\":\"Oliver Stone\"}\n",
    "  - slot{\"director\":\"Oliver Stone\"}\n",
    "  - utter_movie_search_by_director\n",
    "  - action_movie_search_by_director\n",
    "* search_movie_actors{\"movie_title\":\"Fight Club\"}\n",
    "  - slot{\"movie_title\":\"Fight Club\"}\n",
    "  - utter_search_movie_actors\n",
    "  - action_search_movie_actors  \n",
    "* bye\n",
    "  - utter_bye\n",
    "    \n",
    "## happy path 2\n",
    "* greet\n",
    "  - utter_greet\n",
    "* get_introduction\n",
    "  - utter_introduction\n",
    "* movie_search_by_actor{\"actor\":\"Brad Pitt\"}\n",
    "  - slot{\"actor\":\"Brad Pitt\"}\n",
    "  - utter_movie_search_by_actor\n",
    "  - action_movie_search_by_actor\n",
    "* release_year_search_by_movie{\"movie_title\":\"Fight Club\"}\n",
    "  - slot{\"movie_title\":\"Fight Club\"}\n",
    "  - utter_search_release_year_for_movie\n",
    "  - action_search_release_year_for_movie\n",
    "* search_rating_for_movie{\"movie_title\":\"Fight Club\"}\n",
    "  - slot{\"movie_title\":\"Fight Club\"}\n",
    "  - utter_search_rating_for_movie\n",
    "  - action_search_rating_for_movie\n",
    "\n",
    "## happy path 3\n",
    "* greet\n",
    "  - utter_greet\n",
    "* get_introduction\n",
    "  - utter_introduction\n",
    "* movie_search_by_director{\"director\":\"Ridley Scott\"}\n",
    "  - slot{\"director\":\"Ridley Scott\"}\n",
    "  - utter_movie_search_by_director\n",
    "  - action_movie_search_by_director\n",
    "* release_year_search_by_movie{\"movie_title\":\"Alien\"}\n",
    "  - slot{\"movie_title\":\"Alien\"}\n",
    "  - utter_search_release_year_for_movie\n",
    "  - action_search_release_year_for_movie\n",
    "* search_overview_for_movie{\"movie_title\":\"Alien\"}\n",
    "  - slot{\"movie_title\":\"Alien\"}\n",
    "  - utter_search_overview_for_movie\n",
    "  - action_search_overview_for_movie\n",
    "* search_rating_for_movie{\"movie_title\":\"Alien\"}\n",
    "  - slot{\"movie_title\":\"Alien\"}\n",
    "  - utter_search_rating_for_movie\n",
    "  - action_search_rating_for_movie\n",
    "\n",
    "## happy path 4\n",
    "* greet\n",
    "  - utter_greet\n",
    "* get_introduction\n",
    "  - utter_introduction\n",
    "* movie_search_by_director{\"director\":\"Stanley Kubrick\"}\n",
    "  - slot{\"director\":\"Stanley Kubrick\"}\n",
    "  - utter_movie_search_by_director\n",
    "  - action_movie_search_by_director\n",
    "* search_overview_for_movie{\"movie_title\":\"The Shining\"}\n",
    "  - slot{\"movie_title\":\"The Shining\"}\n",
    "  - utter_search_overview_for_movie\n",
    "  - action_search_overview_for_movie\n",
    "* release_year_search_by_movie\n",
    "  - utter_search_release_year_for_movie\n",
    "  - action_search_release_year_for_movie\n",
    "* search_rating_for_movie{\"movie_title\":\"The Shining\"}\n",
    "  - slot{\"movie_title\":\"The Shining\"}\n",
    "  - utter_search_rating_for_movie\n",
    "  - action_search_rating_for_movie\n",
    "\n",
    "## happy path 5\n",
    "* greet\n",
    "  - utter_greet\n",
    "* get_introduction\n",
    "  - utter_introduction\n",
    "* movie_search_by_actor{\"actor\":\"Jennifer Lawrence\"}\n",
    "  - slot{\"actor\":\"Jennifer Lawrence\"}\n",
    "  - utter_movie_search_by_actor\n",
    "  - action_movie_search_by_actor\n",
    "* movie_search_by_actor{\"actor\":\"Sting\"}\n",
    "  - slot{\"actor\":\"Sting\"}\n",
    "  - utter_movie_search_by_actor\n",
    "  - action_movie_search_by_actor\n",
    "* movie_search_by_actor{\"actor\":\"Brad Pitt\"}\n",
    "  - slot{\"actor\":\"Brad Pitt\"}\n",
    "  - utter_movie_search_by_actor\n",
    "  - action_movie_search_by_actor\n",
    "* search_overview_for_movie{\"movie_title\":\"Se7en\"}\n",
    "  - slot{\"movie_title\":\"Se7en\"}\n",
    "  - utter_search_overview_for_movie\n",
    "  - action_search_overview_for_movie\n",
    "* release_year_search_by_movie{\"movie_title\":\"Se7en\"}\n",
    "  - slot{\"movie_title\":\"Se7en\"}\n",
    "  - utter_search_release_year_for_movie\n",
    "  - action_search_release_year_for_movie\n",
    "  \n",
    "## happy path 6\n",
    "* greet\n",
    "    - utter_greet\n",
    "* get_introduction\n",
    "    - utter_introduction\n",
    "* search_overview_for_movie{\"movie_title\":\"alien\"}\n",
    "    - slot{\"movie_title\":\"alien\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "* search_popularity_by_year{\"release\":\"2019\"}\n",
    "    - slot{\"release\":\"2019\"}\n",
    "    - utter_search_popularity_by_year\n",
    "    - action_search_popularity_by_year\n",
    "* release_year_search_by_movie{\"movie_title\":\"Joker\"}\n",
    "    - slot{\"movie_title\":\"Joker\"}\n",
    "    - utter_search_release_year_for_movie\n",
    "    - action_search_release_year_for_movie\n",
    "* search_movie_director{\"movie_title\":\"Joker\"}\n",
    "    - slot{\"movie_title\":\"Joker\"}\n",
    "    - utter_search_movie_director\n",
    "    - action_search_movie_director\n",
    "* search_overview_for_movie\n",
    "    - slot{\"movie_title\":\"Joker\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "    \n",
    "## happy path 7\n",
    "* search_popularity_by_year{\"release\":\"2019\"}\n",
    "    - slot{\"release\":\"2019\"}\n",
    "    - utter_search_popularity_by_year\n",
    "    - action_search_popularity_by_year\n",
    "* search_movie_director{\"movie_title\":\"Logan\"}\n",
    "    - slot{\"movie_title\":\"Logan\"}\n",
    "    - utter_search_movie_director\n",
    "    - action_search_movie_director\n",
    "* search_rating_for_movie{\"movie_title\":\"Logan\"}\n",
    "  - slot{\"movie_title\":\"Logan\"}\n",
    "  - utter_search_rating_for_movie\n",
    "  - action_search_rating_for_movie\n",
    "* search_overview_for_movie{\"movie_title\":\"Logan\"}\n",
    "    - slot{\"movie_title\":\"Logan\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "* release_year_search_by_movie{\"movie_title\":\"Logan\"}\n",
    "    - slot{\"movie_title\":\"Logan\"}\n",
    "    - utter_search_release_year_for_movie\n",
    "    - action_search_release_year_for_movie\n",
    "    \n",
    "## happy path 8\n",
    "* greet\n",
    "    - utter_greet\n",
    "* get_introduction\n",
    "    - utter_introduction\n",
    "* search_popularity_by_year{\"release\":\"1994\"}\n",
    "    - slot{\"release\":\"1994\"}\n",
    "    - utter_search_popularity_by_year\n",
    "    - action_search_popularity_by_year\n",
    "* search_overview_for_movie{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - slot{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "* search_movie_actors{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - slot{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - utter_search_movie_actors\n",
    "    - action_search_movie_actors\n",
    "* search_movie_director{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - slot{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - utter_search_movie_director\n",
    "    - action_search_movie_director\n",
    "\n",
    "## happy path 9\n",
    "* greet\n",
    "    - utter_greet\n",
    "* get_introduction\n",
    "    - utter_introduction\n",
    "* search_popularity_by_year{\"release\":\"1994\"}\n",
    "    - slot{\"release\":\"1994\"}\n",
    "    - utter_search_popularity_by_year\n",
    "    - action_search_popularity_by_year\n",
    "* search_overview_for_movie{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - slot{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "* search_movie_actors{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - slot{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - utter_search_movie_actors\n",
    "    - action_search_movie_actors\n",
    "* search_movie_director{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - slot{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - utter_search_movie_director{\"movie_title\":\"Pulp Fiction\"}\n",
    "    - action_search_movie_director\n",
    "* search_popularity_by_year{\"release\":\"2018\"}\n",
    "    - slot{\"release\":\"2018\"}\n",
    "    - utter_search_popularity_by_year\n",
    "    - action_search_popularity_by_year\n",
    "* search_overview_for_movie{\"movie_title\":\"Black Panther\"}\n",
    "    - slot{\"movie_title\":\"Black Panther\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "* release_year_search_by_movie{\"movie_title\":\"Black Panther\"}\n",
    "    - slot{\"movie_title\":\"Black Panther\"}\n",
    "    - utter_search_release_year_for_movie\n",
    "    - action_search_release_year_for_movie\n",
    "    \n",
    "## happy path 10\n",
    "* search_popularity_by_year{\"release\":\"2007\"}\n",
    "    - slot{\"release\":\"2007\"}\n",
    "    - utter_search_popularity_by_year\n",
    "    - action_search_popularity_by_year\n",
    "* search_overview_for_movie{\"movie_title\":\"Coco\"}\n",
    "    - slot{\"movie_title\":\"Coco\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "* search_movie_director{\"movie_title\":\"Coco\"}\n",
    "    - slot{\"movie_title\":\"Coco\"}\n",
    "    - utter_search_movie_director{\"movie_title\":\"Coco\"}\n",
    "    - action_search_movie_director\n",
    "* search_movie_actors{\"movie_title\":\"Coco\"}\n",
    "    - slot{\"movie_title\":\"Coco\"}\n",
    "    - utter_search_movie_actors\n",
    "    - action_search_movie_actors\n",
    "* search_similar_movies{\"movie_title\":\"Coco\"}\n",
    "    - slot{\"movie_title\":\"Coco\"}\n",
    "    - utter_search_similar_movies\n",
    "    - action_search_similar_movies \n",
    "    \n",
    "## happy path 11\n",
    "* search_overview_for_movie{\"movie_title\":\"Coco\"}\n",
    "    - slot{\"movie_title\":\"Coco\"}\n",
    "    - utter_search_overview_for_movie\n",
    "    - action_search_overview_for_movie\n",
    "* search_similar_movies{\"movie_title\":\"Coco\"}\n",
    "    - slot{\"movie_title\":\"Coco\"}\n",
    "    - utter_search_similar_movies\n",
    "    - action_search_similar_movies\n",
    "* search_movie_actors{\"movie_title\":\"Coco\"}\n",
    "    - slot{\"movie_title\":\"Coco\"}\n",
    "    - utter_search_movie_actors\n",
    "    - action_search_movie_actors\n",
    "\n",
    "\"\"\"\n",
    "write_to_file(movichtr_training_file_stories, movichtr_stories)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "movichtr_domain = \"\"\"\n",
    "intents:\n",
    "  - greet\n",
    "  - goodbye\n",
    "  - affirm\n",
    "  - deny\n",
    "  - bye\n",
    "  - movie_search_by_actor\n",
    "  - movie_search_by_director\n",
    "  - release_year_search_by_movie\n",
    "  - search_overview_for_movie\n",
    "  - search_rating_for_movie\n",
    "  - search_popularity_by_year\n",
    "  - search_movie_director\n",
    "  - search_movie_actors\n",
    "  - search_similar_movies\n",
    "  - get_introduction\n",
    "\n",
    "actions:\n",
    "- utter_greet\n",
    "- utter_cheer_up\n",
    "- utter_did_that_help\n",
    "- utter_happy\n",
    "- utter_bye\n",
    "- action_movie_search_by_actor\n",
    "- utter_movie_search_by_actor\n",
    "- action_movie_search_by_director\n",
    "- utter_movie_search_by_director\n",
    "- action_search_release_year_for_movie\n",
    "- utter_search_release_year_for_movie\n",
    "- action_search_overview_for_movie\n",
    "- utter_search_overview_for_movie\n",
    "- action_search_rating_for_movie\n",
    "- utter_search_rating_for_movie\n",
    "- action_search_popularity_by_year\n",
    "- utter_search_popularity_by_year\n",
    "- action_search_movie_director\n",
    "- utter_search_movie_director\n",
    "- action_search_movie_actors\n",
    "- utter_search_movie_actors\n",
    "- action_search_similar_movies\n",
    "- utter_search_similar_movies\n",
    "- utter_introduction\n",
    "\n",
    "entities:\n",
    "- movie_title\n",
    "- director\n",
    "- actor\n",
    "- release\n",
    "\n",
    "\n",
    "slots:\n",
    "    movie_title:\n",
    "        type: text\n",
    "    director:\n",
    "        type: text\n",
    "    actor:\n",
    "        type: text\n",
    "    release:\n",
    "        type: text\n",
    "        \n",
    "templates:\n",
    "  utter_greet:\n",
    "  - text: \"Hey! How are you?\"\n",
    "\n",
    "  utter_cheer_up:\n",
    "  - text: \"Here is something to cheer you up:\"\n",
    "    image: \"https://i.imgur.com/nGF1K8f.jpg\"\n",
    "\n",
    "  utter_did_that_help:\n",
    "  - text: \"Did that help you?\"\n",
    "\n",
    "  utter_happy:\n",
    "  - text: \"Great, carry on!\"\n",
    "  \n",
    "  utter_movie_search_by_actor:\n",
    "  - text: \"I'm ready to search for movies with {actor}\"\n",
    "  \n",
    "  utter_movie_search_by_director:\n",
    "  - text: \"Let me find movies directed by {director}\"\n",
    "  \n",
    "  utter_search_release_year_for_movie:\n",
    "  - text: \"I'm checking when {movie_title} was released\"\n",
    "  \n",
    "  utter_search_overview_for_movie:\n",
    "  - text: \"Looking for the plot of {movie_title}\"\n",
    "    \n",
    "  utter_search_rating_for_movie:\n",
    "  - text: \"Looking for the rating of {movie_title}\"\n",
    "  \n",
    "  utter_search_popularity_by_year:\n",
    "  - text: \"Looking for popular movies from {release}\"\n",
    "  \n",
    "  utter_search_movie_director:\n",
    "  - text: \"Looking for the director of {movie_title}\"\n",
    "  \n",
    "  utter_search_movie_actors:\n",
    "  - text: \"Looking for who acted in {movie_title}\"\n",
    "  \n",
    "  utter_search_similar_movies:\n",
    "  - text: \"Looking for similar movies to {movie_title}\"\n",
    "  \n",
    "  utter_introduction:\n",
    "  - text: \"hi! I am MoviCHTR. I can search movies by director, actor, get you a release date or rating using a chosen title. I am even trying to learn movie popularity based on a particular year\"\n",
    "   \n",
    "  utter_bye:\n",
    "  - text: \"Bye\"\n",
    "\n",
    "  \n",
    "\"\"\"\n",
    "write_to_file(movichtr_domain_path, movichtr_domain)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To prepare the bot to be deployed on slack I will open an option for direct editting from jupyter:\n",
    "\n",
    "### I also made a guide considering that it has quite a few steps and I got stuck many times: https://docs.google.com/document/d/1-qo1qt7wI0FvspyBZTzb7oUx4-NeAlf-Cy-WHokxHPA/edit?usp=sharing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "credentials = \"\"\"\n",
    "# This file contains the credentials for the voice & chat platforms\n",
    "# which your bot is using.\n",
    "# https://rasa.com/docs/rasa/user-guide/messaging-and-voice-channels/\n",
    "\n",
    "rest:\n",
    "#  # you don't need to provide anything here - this channel doesn't\n",
    "#  # require any credentials\n",
    "\n",
    "\n",
    "#facebook:\n",
    "#  verify: \"<verify>\"\n",
    "#  secret: \"<your secret>\"\n",
    "#  page-access-token: \"<your page access token>\"\n",
    "\n",
    "slack:\n",
    " slack_token: \"xoxb-720064421394-807667952307-WjHRR4bMmBeTtpYD2QlQIQT9\"\n",
    " slack_channel: \"#challenge-bot-testing\"\n",
    "\n",
    "#socketio:\n",
    "#  user_message_evt: <event name for user message>\n",
    "#  bot_message_evt: <event name for but messages>\n",
    "#  session_persistence: <true/false>\n",
    "\n",
    "#mattermost:\n",
    "#  url: \"https://<mattermost instance>/api/v4\"\n",
    "#  team: \"<mattermost team>\"\n",
    "#  user: \"<bot username>\"\n",
    "#  pw: \"<bot token>\"\n",
    "#  webhook_url: \"<callback URL>\"\n",
    "\n",
    "rasa:\n",
    "\n",
    "  url: \"http://localhost:5002/api\"  \n",
    "\"\"\"\n",
    "write_to_file(movichtr_credentials_path, credentials)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Once we have stories and domains defined and enriched we can train Rasa Core"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_core():\n",
    "    agent = Agent(movichtr_domain_path, policies=[MemoizationPolicy(), KerasPolicy(), MappingPolicy()])\n",
    "   \n",
    "    loop = asyncio.get_event_loop()\n",
    "    training_data = loop.run_until_complete(agent.load_data(movichtr_training_files + 'stories.md'))\n",
    "    \n",
    "    agent.train(training_data)\n",
    "    \n",
    "    agent.persist(movichtr_models_output + movichtr_model_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Processed Story Blocks: 100%|███████████████████████████████████████████| 24/24 [00:00<00:00, 460.29it/s, # trackers=1]\n",
      "Processed Story Blocks: 100%|███████████████████████████████████████████| 24/24 [00:00<00:00, 43.27it/s, # trackers=23]\n",
      "Processed Story Blocks: 100%|███████████████████████████████████████████| 24/24 [00:00<00:00, 36.23it/s, # trackers=39]\n",
      "Processed Story Blocks: 100%|███████████████████████████████████████████| 24/24 [00:00<00:00, 27.82it/s, # trackers=44]\n",
      "Processed trackers: 100%|███████████████████████████████████████████████| 24/24 [00:00<00:00, 65.81it/s, # actions=150]\n",
      "Processed actions: 150it [00:00, 591.52it/s, # examples=150]\n",
      "Processed trackers: 100%|████████████████████████████████████████████| 524/524 [00:24<00:00, 21.35it/s, # actions=1616]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "masking (Masking)            (None, 5, 55)             0         \n",
      "_________________________________________________________________\n",
      "lstm (LSTM)                  (None, 32)                11264     \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 32)                1056      \n",
      "_________________________________________________________________\n",
      "activation (Activation)      (None, 32)                0         \n",
      "=================================================================\n",
      "Total params: 12,320\n",
      "Trainable params: 12,320\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "Epoch 1/100\n",
      "1616/1616 [==============================] - ETA: 41s - loss: 3.4401 - acc: 0.03 - ETA: 4s - loss: 3.3991 - acc: 0.1328 - ETA: 2s - loss: 3.3446 - acc: 0.189 - ETA: 1s - loss: 3.3092 - acc: 0.220 - ETA: 0s - loss: 3.2707 - acc: 0.247 - ETA: 0s - loss: 3.2242 - acc: 0.262 - ETA: 0s - loss: 3.1750 - acc: 0.274 - 1s 734us/sample - loss: 3.1628 - acc: 0.2760\n",
      "Epoch 2/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 2.7695 - acc: 0.343 - ETA: 0s - loss: 2.7602 - acc: 0.378 - ETA: 0s - loss: 2.7223 - acc: 0.382 - ETA: 0s - loss: 2.7323 - acc: 0.358 - ETA: 0s - loss: 2.7325 - acc: 0.348 - ETA: 0s - loss: 2.7216 - acc: 0.348 - 0s 193us/sample - loss: 2.7191 - acc: 0.3422\n",
      "Epoch 3/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 2.5090 - acc: 0.343 - ETA: 0s - loss: 2.5783 - acc: 0.354 - ETA: 0s - loss: 2.5005 - acc: 0.376 - ETA: 0s - loss: 2.5138 - acc: 0.363 - ETA: 0s - loss: 2.5364 - acc: 0.350 - ETA: 0s - loss: 2.5451 - acc: 0.346 - ETA: 0s - loss: 2.5394 - acc: 0.346 - 0s 213us/sample - loss: 2.5458 - acc: 0.3422\n",
      "Epoch 4/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 2.3746 - acc: 0.343 - ETA: 0s - loss: 2.4086 - acc: 0.363 - ETA: 0s - loss: 2.3634 - acc: 0.381 - ETA: 0s - loss: 2.4090 - acc: 0.359 - ETA: 0s - loss: 2.4319 - acc: 0.348 - ETA: 0s - loss: 2.4258 - acc: 0.348 - 0s 191us/sample - loss: 2.4347 - acc: 0.3422\n",
      "Epoch 5/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 2.2786 - acc: 0.343 - ETA: 0s - loss: 2.2969 - acc: 0.363 - ETA: 0s - loss: 2.2478 - acc: 0.381 - ETA: 0s - loss: 2.3095 - acc: 0.354 - ETA: 0s - loss: 2.3159 - acc: 0.346 - ETA: 0s - loss: 2.3109 - acc: 0.346 - 0s 180us/sample - loss: 2.3150 - acc: 0.3428\n",
      "Epoch 6/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 2.1362 - acc: 0.343 - ETA: 0s - loss: 2.1688 - acc: 0.363 - ETA: 0s - loss: 2.1251 - acc: 0.381 - ETA: 0s - loss: 2.1819 - acc: 0.354 - ETA: 0s - loss: 2.1811 - acc: 0.347 - ETA: 0s - loss: 2.1868 - acc: 0.343 - 0s 177us/sample - loss: 2.1818 - acc: 0.3428\n",
      "Epoch 7/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 2.0486 - acc: 0.375 - ETA: 0s - loss: 1.9908 - acc: 0.393 - ETA: 0s - loss: 1.9930 - acc: 0.388 - ETA: 0s - loss: 2.0302 - acc: 0.366 - ETA: 0s - loss: 2.0310 - acc: 0.363 - ETA: 0s - loss: 2.0254 - acc: 0.362 - 0s 198us/sample - loss: 2.0289 - acc: 0.3595\n",
      "Epoch 8/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 1.8475 - acc: 0.406 - ETA: 0s - loss: 1.8518 - acc: 0.430 - ETA: 0s - loss: 1.8163 - acc: 0.427 - ETA: 0s - loss: 1.8445 - acc: 0.405 - ETA: 0s - loss: 1.8657 - acc: 0.400 - ETA: 0s - loss: 1.8575 - acc: 0.399 - 0s 193us/sample - loss: 1.8644 - acc: 0.3967\n",
      "Epoch 9/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 1.6838 - acc: 0.500 - ETA: 0s - loss: 1.6426 - acc: 0.512 - ETA: 0s - loss: 1.6323 - acc: 0.495 - ETA: 0s - loss: 1.6697 - acc: 0.473 - ETA: 0s - loss: 1.6723 - acc: 0.476 - ETA: 0s - loss: 1.6690 - acc: 0.484 - 0s 187us/sample - loss: 1.6741 - acc: 0.4845\n",
      "Epoch 10/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 1.5555 - acc: 0.562 - ETA: 0s - loss: 1.5160 - acc: 0.579 - ETA: 0s - loss: 1.4749 - acc: 0.578 - ETA: 0s - loss: 1.5089 - acc: 0.558 - ETA: 0s - loss: 1.4972 - acc: 0.573 - ETA: 0s - loss: 1.4949 - acc: 0.572 - 0s 185us/sample - loss: 1.4944 - acc: 0.5749\n",
      "Epoch 11/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 1.4320 - acc: 0.656 - ETA: 0s - loss: 1.3683 - acc: 0.635 - ETA: 0s - loss: 1.3085 - acc: 0.650 - ETA: 0s - loss: 1.3277 - acc: 0.647 - ETA: 0s - loss: 1.3328 - acc: 0.648 - ETA: 0s - loss: 1.3215 - acc: 0.657 - 0s 200us/sample - loss: 1.3219 - acc: 0.6597\n",
      "Epoch 12/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 1.1630 - acc: 0.781 - ETA: 0s - loss: 1.1462 - acc: 0.735 - ETA: 0s - loss: 1.1161 - acc: 0.737 - ETA: 0s - loss: 1.1472 - acc: 0.720 - ETA: 0s - loss: 1.1461 - acc: 0.727 - ETA: 0s - loss: 1.1367 - acc: 0.729 - 0s 185us/sample - loss: 1.1397 - acc: 0.7314\n",
      "Epoch 13/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 1.0551 - acc: 0.750 - ETA: 0s - loss: 0.9905 - acc: 0.793 - ETA: 0s - loss: 0.9792 - acc: 0.778 - ETA: 0s - loss: 1.0149 - acc: 0.771 - ETA: 0s - loss: 1.0134 - acc: 0.778 - ETA: 0s - loss: 0.9936 - acc: 0.790 - 0s 178us/sample - loss: 0.9939 - acc: 0.7915\n",
      "Epoch 14/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.9445 - acc: 0.781 - ETA: 0s - loss: 0.8404 - acc: 0.856 - ETA: 0s - loss: 0.8104 - acc: 0.860 - ETA: 0s - loss: 0.8279 - acc: 0.861 - ETA: 0s - loss: 0.8287 - acc: 0.863 - ETA: 0s - loss: 0.8231 - acc: 0.863 - 0s 187us/sample - loss: 0.8306 - acc: 0.8651\n",
      "Epoch 15/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.7319 - acc: 0.875 - ETA: 0s - loss: 0.7174 - acc: 0.900 - ETA: 0s - loss: 0.7036 - acc: 0.895 - ETA: 0s - loss: 0.7288 - acc: 0.891 - ETA: 0s - loss: 0.7243 - acc: 0.894 - ETA: 0s - loss: 0.7267 - acc: 0.892 - 0s 174us/sample - loss: 0.7273 - acc: 0.8917\n",
      "Epoch 16/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.6390 - acc: 0.937 - ETA: 0s - loss: 0.6529 - acc: 0.900 - ETA: 0s - loss: 0.6146 - acc: 0.910 - ETA: 0s - loss: 0.6378 - acc: 0.910 - ETA: 0s - loss: 0.6359 - acc: 0.914 - ETA: 0s - loss: 0.6272 - acc: 0.920 - 0s 178us/sample - loss: 0.6289 - acc: 0.9202\n",
      "Epoch 17/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.6554 - acc: 0.906 - ETA: 0s - loss: 0.5724 - acc: 0.909 - ETA: 0s - loss: 0.5515 - acc: 0.921 - ETA: 0s - loss: 0.5774 - acc: 0.913 - ETA: 0s - loss: 0.5691 - acc: 0.919 - ETA: 0s - loss: 0.5654 - acc: 0.919 - 0s 190us/sample - loss: 0.5711 - acc: 0.9196\n",
      "Epoch 18/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.4885 - acc: 1.000 - ETA: 0s - loss: 0.4905 - acc: 0.943 - ETA: 0s - loss: 0.4616 - acc: 0.951 - ETA: 0s - loss: 0.4709 - acc: 0.947 - ETA: 0s - loss: 0.4703 - acc: 0.946 - ETA: 0s - loss: 0.4657 - acc: 0.946 - 0s 183us/sample - loss: 0.4676 - acc: 0.9480\n",
      "Epoch 19/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.3630 - acc: 0.968 - ETA: 0s - loss: 0.4354 - acc: 0.960 - ETA: 0s - loss: 0.4030 - acc: 0.957 - ETA: 0s - loss: 0.4200 - acc: 0.956 - ETA: 0s - loss: 0.4158 - acc: 0.957 - ETA: 0s - loss: 0.4144 - acc: 0.956 - 0s 183us/sample - loss: 0.4126 - acc: 0.9567\n",
      "Epoch 20/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.3601 - acc: 0.968 - ETA: 0s - loss: 0.3889 - acc: 0.951 - ETA: 0s - loss: 0.3724 - acc: 0.954 - ETA: 0s - loss: 0.3671 - acc: 0.962 - ETA: 0s - loss: 0.3773 - acc: 0.962 - ETA: 0s - loss: 0.3804 - acc: 0.957 - 0s 183us/sample - loss: 0.3847 - acc: 0.9585\n",
      "Epoch 21/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.3201 - acc: 0.968 - ETA: 0s - loss: 0.3128 - acc: 0.965 - ETA: 0s - loss: 0.3005 - acc: 0.973 - ETA: 0s - loss: 0.3189 - acc: 0.967 - ETA: 0s - loss: 0.3313 - acc: 0.964 - ETA: 0s - loss: 0.3351 - acc: 0.963 - 0s 184us/sample - loss: 0.3339 - acc: 0.9641\n",
      "Epoch 22/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.2990 - acc: 1.000 - ETA: 0s - loss: 0.2621 - acc: 0.978 - ETA: 0s - loss: 0.2743 - acc: 0.975 - ETA: 0s - loss: 0.2781 - acc: 0.972 - ETA: 0s - loss: 0.2836 - acc: 0.968 - ETA: 0s - loss: 0.2766 - acc: 0.971 - 0s 178us/sample - loss: 0.2763 - acc: 0.9722\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 23/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.3080 - acc: 0.906 - ETA: 0s - loss: 0.2661 - acc: 0.965 - ETA: 0s - loss: 0.2695 - acc: 0.963 - ETA: 0s - loss: 0.2644 - acc: 0.962 - ETA: 0s - loss: 0.2663 - acc: 0.960 - ETA: 0s - loss: 0.2600 - acc: 0.963 - ETA: 0s - loss: 0.2631 - acc: 0.964 - 0s 214us/sample - loss: 0.2631 - acc: 0.9635\n",
      "Epoch 24/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.2998 - acc: 0.968 - ETA: 0s - loss: 0.2277 - acc: 0.978 - ETA: 0s - loss: 0.2173 - acc: 0.985 - ETA: 0s - loss: 0.2119 - acc: 0.982 - ETA: 0s - loss: 0.2143 - acc: 0.979 - ETA: 0s - loss: 0.2176 - acc: 0.980 - ETA: 0s - loss: 0.2193 - acc: 0.977 - ETA: 0s - loss: 0.2149 - acc: 0.978 - 0s 249us/sample - loss: 0.2158 - acc: 0.9783\n",
      "Epoch 25/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.2425 - acc: 0.968 - ETA: 0s - loss: 0.1977 - acc: 0.979 - ETA: 0s - loss: 0.1794 - acc: 0.983 - ETA: 0s - loss: 0.1794 - acc: 0.983 - ETA: 0s - loss: 0.1817 - acc: 0.978 - ETA: 0s - loss: 0.1797 - acc: 0.978 - ETA: 0s - loss: 0.1857 - acc: 0.978 - 0s 243us/sample - loss: 0.1913 - acc: 0.9783\n",
      "Epoch 26/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1364 - acc: 0.968 - ETA: 0s - loss: 0.1958 - acc: 0.975 - ETA: 0s - loss: 0.1934 - acc: 0.975 - ETA: 0s - loss: 0.1862 - acc: 0.976 - ETA: 0s - loss: 0.1994 - acc: 0.970 - ETA: 0s - loss: 0.2017 - acc: 0.972 - ETA: 0s - loss: 0.1957 - acc: 0.973 - 0s 223us/sample - loss: 0.1932 - acc: 0.9734\n",
      "Epoch 27/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0870 - acc: 1.000 - ETA: 0s - loss: 0.1251 - acc: 0.993 - ETA: 0s - loss: 0.1543 - acc: 0.984 - ETA: 0s - loss: 0.1553 - acc: 0.986 - ETA: 0s - loss: 0.1590 - acc: 0.984 - ETA: 0s - loss: 0.1663 - acc: 0.983 - ETA: 0s - loss: 0.1730 - acc: 0.980 - 0s 225us/sample - loss: 0.1711 - acc: 0.9802\n",
      "Epoch 28/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1272 - acc: 1.000 - ETA: 0s - loss: 0.1667 - acc: 0.972 - ETA: 0s - loss: 0.1512 - acc: 0.979 - ETA: 0s - loss: 0.1399 - acc: 0.981 - ETA: 0s - loss: 0.1421 - acc: 0.981 - ETA: 0s - loss: 0.1402 - acc: 0.982 - ETA: 0s - loss: 0.1462 - acc: 0.982 - 0s 216us/sample - loss: 0.1448 - acc: 0.9827\n",
      "Epoch 29/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1318 - acc: 1.000 - ETA: 0s - loss: 0.1530 - acc: 0.982 - ETA: 0s - loss: 0.1359 - acc: 0.979 - ETA: 0s - loss: 0.1325 - acc: 0.985 - ETA: 0s - loss: 0.1303 - acc: 0.988 - ETA: 0s - loss: 0.1375 - acc: 0.988 - ETA: 0s - loss: 0.1392 - acc: 0.986 - 0s 223us/sample - loss: 0.1367 - acc: 0.9876\n",
      "Epoch 30/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1318 - acc: 1.000 - ETA: 0s - loss: 0.1236 - acc: 0.989 - ETA: 0s - loss: 0.1144 - acc: 0.989 - ETA: 0s - loss: 0.1145 - acc: 0.988 - ETA: 0s - loss: 0.1131 - acc: 0.988 - ETA: 0s - loss: 0.1255 - acc: 0.984 - ETA: 0s - loss: 0.1246 - acc: 0.984 - 0s 223us/sample - loss: 0.1326 - acc: 0.9833\n",
      "Epoch 31/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1430 - acc: 0.968 - ETA: 0s - loss: 0.1223 - acc: 0.989 - ETA: 0s - loss: 0.1124 - acc: 0.989 - ETA: 0s - loss: 0.1074 - acc: 0.990 - ETA: 0s - loss: 0.1089 - acc: 0.988 - ETA: 0s - loss: 0.1078 - acc: 0.988 - ETA: 0s - loss: 0.1057 - acc: 0.987 - 0s 222us/sample - loss: 0.1117 - acc: 0.9864\n",
      "Epoch 32/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.2515 - acc: 0.968 - ETA: 0s - loss: 0.1197 - acc: 0.984 - ETA: 0s - loss: 0.1321 - acc: 0.979 - ETA: 0s - loss: 0.1230 - acc: 0.981 - ETA: 0s - loss: 0.1240 - acc: 0.982 - ETA: 0s - loss: 0.1269 - acc: 0.984 - ETA: 0s - loss: 0.1231 - acc: 0.984 - 0s 214us/sample - loss: 0.1249 - acc: 0.9839\n",
      "Epoch 33/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0759 - acc: 1.000 - ETA: 0s - loss: 0.0972 - acc: 0.979 - ETA: 0s - loss: 0.1170 - acc: 0.979 - ETA: 0s - loss: 0.1138 - acc: 0.981 - ETA: 0s - loss: 0.1072 - acc: 0.984 - ETA: 0s - loss: 0.1058 - acc: 0.984 - ETA: 0s - loss: 0.1094 - acc: 0.982 - 0s 216us/sample - loss: 0.1100 - acc: 0.9827\n",
      "Epoch 34/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1441 - acc: 0.968 - ETA: 0s - loss: 0.1067 - acc: 0.993 - ETA: 0s - loss: 0.0924 - acc: 0.992 - ETA: 0s - loss: 0.0926 - acc: 0.992 - ETA: 0s - loss: 0.0978 - acc: 0.991 - ETA: 0s - loss: 0.1031 - acc: 0.987 - ETA: 0s - loss: 0.1086 - acc: 0.987 - 0s 217us/sample - loss: 0.1088 - acc: 0.9876\n",
      "Epoch 35/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1984 - acc: 1.000 - ETA: 0s - loss: 0.0816 - acc: 0.993 - ETA: 0s - loss: 0.0832 - acc: 0.992 - ETA: 0s - loss: 0.0845 - acc: 0.990 - ETA: 0s - loss: 0.0891 - acc: 0.987 - ETA: 0s - loss: 0.0954 - acc: 0.986 - ETA: 0s - loss: 0.0939 - acc: 0.986 - 0s 230us/sample - loss: 0.0948 - acc: 0.9864\n",
      "Epoch 36/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1307 - acc: 0.968 - ETA: 0s - loss: 0.0999 - acc: 0.979 - ETA: 0s - loss: 0.0958 - acc: 0.979 - ETA: 0s - loss: 0.1058 - acc: 0.976 - ETA: 0s - loss: 0.1017 - acc: 0.981 - ETA: 0s - loss: 0.1075 - acc: 0.982 - ETA: 0s - loss: 0.1065 - acc: 0.980 - 0s 219us/sample - loss: 0.1040 - acc: 0.9814\n",
      "Epoch 37/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0925 - acc: 1.000 - ETA: 0s - loss: 0.0637 - acc: 1.000 - ETA: 0s - loss: 0.0783 - acc: 0.993 - ETA: 0s - loss: 0.0738 - acc: 0.992 - ETA: 0s - loss: 0.0761 - acc: 0.990 - ETA: 0s - loss: 0.0807 - acc: 0.989 - ETA: 0s - loss: 0.0815 - acc: 0.990 - 0s 221us/sample - loss: 0.0820 - acc: 0.9901\n",
      "Epoch 38/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1088 - acc: 0.968 - ETA: 0s - loss: 0.0673 - acc: 0.991 - ETA: 0s - loss: 0.0877 - acc: 0.983 - ETA: 0s - loss: 0.0864 - acc: 0.983 - ETA: 0s - loss: 0.0826 - acc: 0.983 - ETA: 0s - loss: 0.0797 - acc: 0.986 - ETA: 0s - loss: 0.0795 - acc: 0.986 - 0s 231us/sample - loss: 0.0807 - acc: 0.9870\n",
      "Epoch 39/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0523 - acc: 1.000 - ETA: 0s - loss: 0.0830 - acc: 0.996 - ETA: 0s - loss: 0.0618 - acc: 0.998 - ETA: 0s - loss: 0.0628 - acc: 0.997 - ETA: 0s - loss: 0.0762 - acc: 0.992 - ETA: 0s - loss: 0.0729 - acc: 0.992 - ETA: 0s - loss: 0.0732 - acc: 0.991 - 0s 216us/sample - loss: 0.0731 - acc: 0.9913\n",
      "Epoch 40/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0511 - acc: 1.000 - ETA: 0s - loss: 0.0482 - acc: 0.996 - ETA: 0s - loss: 0.0586 - acc: 0.993 - ETA: 0s - loss: 0.0731 - acc: 0.990 - ETA: 0s - loss: 0.0716 - acc: 0.990 - ETA: 0s - loss: 0.0805 - acc: 0.986 - ETA: 0s - loss: 0.0755 - acc: 0.987 - 0s 225us/sample - loss: 0.0758 - acc: 0.9876\n",
      "Epoch 41/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0127 - acc: 1.000 - ETA: 0s - loss: 0.0721 - acc: 0.989 - ETA: 0s - loss: 0.0665 - acc: 0.987 - ETA: 0s - loss: 0.0715 - acc: 0.986 - ETA: 0s - loss: 0.0761 - acc: 0.985 - ETA: 0s - loss: 0.0778 - acc: 0.985 - ETA: 0s - loss: 0.0786 - acc: 0.984 - 0s 212us/sample - loss: 0.0774 - acc: 0.9851\n",
      "Epoch 42/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0881 - acc: 0.968 - ETA: 0s - loss: 0.0941 - acc: 0.981 - ETA: 0s - loss: 0.0893 - acc: 0.982 - ETA: 0s - loss: 0.0834 - acc: 0.985 - ETA: 0s - loss: 0.0823 - acc: 0.986 - ETA: 0s - loss: 0.0835 - acc: 0.984 - 0s 206us/sample - loss: 0.0821 - acc: 0.9845\n",
      "Epoch 43/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0189 - acc: 1.000 - ETA: 0s - loss: 0.0480 - acc: 0.993 - ETA: 0s - loss: 0.0459 - acc: 0.993 - ETA: 0s - loss: 0.0532 - acc: 0.989 - ETA: 0s - loss: 0.0548 - acc: 0.988 - ETA: 0s - loss: 0.0547 - acc: 0.989 - ETA: 0s - loss: 0.0556 - acc: 0.987 - 0s 231us/sample - loss: 0.0589 - acc: 0.9882\n",
      "Epoch 44/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0188 - acc: 1.000 - ETA: 0s - loss: 0.0625 - acc: 0.987 - ETA: 0s - loss: 0.0636 - acc: 0.987 - ETA: 0s - loss: 0.0642 - acc: 0.989 - ETA: 0s - loss: 0.0737 - acc: 0.988 - ETA: 0s - loss: 0.0742 - acc: 0.988 - ETA: 0s - loss: 0.0725 - acc: 0.989 - 0s 210us/sample - loss: 0.0721 - acc: 0.9895\n",
      "Epoch 45/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0593 - acc: 1.000 - ETA: 0s - loss: 0.0785 - acc: 0.987 - ETA: 0s - loss: 0.0626 - acc: 0.991 - ETA: 0s - loss: 0.0643 - acc: 0.989 - ETA: 0s - loss: 0.0672 - acc: 0.988 - ETA: 0s - loss: 0.0669 - acc: 0.990 - ETA: 0s - loss: 0.0737 - acc: 0.989 - 0s 218us/sample - loss: 0.0749 - acc: 0.9889\n",
      "Epoch 46/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1151 - acc: 0.968 - ETA: 0s - loss: 0.0497 - acc: 0.989 - ETA: 0s - loss: 0.0406 - acc: 0.994 - ETA: 0s - loss: 0.0538 - acc: 0.992 - ETA: 0s - loss: 0.0549 - acc: 0.991 - ETA: 0s - loss: 0.0601 - acc: 0.991 - ETA: 0s - loss: 0.0561 - acc: 0.992 - ETA: 0s - loss: 0.0633 - acc: 0.991 - 0s 253us/sample - loss: 0.0630 - acc: 0.9913\n",
      "Epoch 47/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1244 - acc: 0.937 - ETA: 0s - loss: 0.0757 - acc: 0.986 - ETA: 0s - loss: 0.0715 - acc: 0.988 - ETA: 0s - loss: 0.0710 - acc: 0.989 - ETA: 0s - loss: 0.0669 - acc: 0.990 - ETA: 0s - loss: 0.0702 - acc: 0.988 - ETA: 0s - loss: 0.0657 - acc: 0.989 - 0s 223us/sample - loss: 0.0649 - acc: 0.9889\n",
      "Epoch 48/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0282 - acc: 1.000 - ETA: 0s - loss: 0.0434 - acc: 0.996 - ETA: 0s - loss: 0.0467 - acc: 0.994 - ETA: 0s - loss: 0.0466 - acc: 0.994 - ETA: 0s - loss: 0.0485 - acc: 0.994 - ETA: 0s - loss: 0.0533 - acc: 0.991 - ETA: 0s - loss: 0.0562 - acc: 0.989 - 0s 235us/sample - loss: 0.0552 - acc: 0.9901\n",
      "Epoch 49/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0443 - acc: 1.000 - ETA: 0s - loss: 0.0414 - acc: 0.996 - ETA: 0s - loss: 0.0440 - acc: 0.994 - ETA: 0s - loss: 0.0443 - acc: 0.996 - ETA: 0s - loss: 0.0540 - acc: 0.996 - ETA: 0s - loss: 0.0568 - acc: 0.994 - ETA: 0s - loss: 0.0570 - acc: 0.994 - 0s 215us/sample - loss: 0.0579 - acc: 0.9944\n",
      "Epoch 50/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0468 - acc: 1.000 - ETA: 0s - loss: 0.0404 - acc: 1.000 - ETA: 0s - loss: 0.0513 - acc: 0.992 - ETA: 0s - loss: 0.0499 - acc: 0.993 - ETA: 0s - loss: 0.0464 - acc: 0.995 - ETA: 0s - loss: 0.0491 - acc: 0.994 - ETA: 0s - loss: 0.0485 - acc: 0.994 - 0s 226us/sample - loss: 0.0514 - acc: 0.9932\n",
      "Epoch 51/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0489 - acc: 1.000 - ETA: 0s - loss: 0.0572 - acc: 0.990 - ETA: 0s - loss: 0.0495 - acc: 0.993 - ETA: 0s - loss: 0.0497 - acc: 0.992 - ETA: 0s - loss: 0.0481 - acc: 0.992 - ETA: 0s - loss: 0.0503 - acc: 0.992 - 0s 203us/sample - loss: 0.0515 - acc: 0.9926\n",
      "Epoch 52/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0318 - acc: 1.000 - ETA: 0s - loss: 0.0275 - acc: 1.000 - ETA: 0s - loss: 0.0453 - acc: 0.997 - ETA: 0s - loss: 0.0475 - acc: 0.994 - ETA: 0s - loss: 0.0524 - acc: 0.992 - ETA: 0s - loss: 0.0512 - acc: 0.993 - ETA: 0s - loss: 0.0501 - acc: 0.992 - 0s 241us/sample - loss: 0.0494 - acc: 0.9932\n",
      "Epoch 53/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0846 - acc: 0.968 - ETA: 0s - loss: 0.0448 - acc: 0.993 - ETA: 0s - loss: 0.0420 - acc: 0.993 - ETA: 0s - loss: 0.0442 - acc: 0.992 - ETA: 0s - loss: 0.0470 - acc: 0.992 - ETA: 0s - loss: 0.0496 - acc: 0.990 - ETA: 0s - loss: 0.0497 - acc: 0.991 - 0s 222us/sample - loss: 0.0495 - acc: 0.9913\n",
      "Epoch 54/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0716 - acc: 1.000 - ETA: 0s - loss: 0.0596 - acc: 0.990 - ETA: 0s - loss: 0.0487 - acc: 0.993 - ETA: 0s - loss: 0.0433 - acc: 0.993 - ETA: 0s - loss: 0.0464 - acc: 0.994 - ETA: 0s - loss: 0.0479 - acc: 0.994 - ETA: 0s - loss: 0.0470 - acc: 0.994 - 0s 223us/sample - loss: 0.0472 - acc: 0.9938\n",
      "Epoch 55/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0124 - acc: 1.000 - ETA: 0s - loss: 0.0247 - acc: 0.996 - ETA: 0s - loss: 0.0231 - acc: 0.996 - ETA: 0s - loss: 0.0306 - acc: 0.993 - ETA: 0s - loss: 0.0350 - acc: 0.993 - ETA: 0s - loss: 0.0444 - acc: 0.991 - ETA: 0s - loss: 0.0438 - acc: 0.991 - 0s 214us/sample - loss: 0.0433 - acc: 0.9913\n",
      "Epoch 56/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0312 - acc: 1.000 - ETA: 0s - loss: 0.0477 - acc: 0.989 - ETA: 0s - loss: 0.0357 - acc: 0.994 - ETA: 0s - loss: 0.0420 - acc: 0.991 - ETA: 0s - loss: 0.0406 - acc: 0.993 - ETA: 0s - loss: 0.0392 - acc: 0.994 - ETA: 0s - loss: 0.0384 - acc: 0.994 - 0s 223us/sample - loss: 0.0389 - acc: 0.9944\n",
      "Epoch 57/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0094 - acc: 1.000 - ETA: 0s - loss: 0.0616 - acc: 0.990 - ETA: 0s - loss: 0.0454 - acc: 0.994 - ETA: 0s - loss: 0.0449 - acc: 0.992 - ETA: 0s - loss: 0.0477 - acc: 0.991 - ETA: 0s - loss: 0.0495 - acc: 0.992 - ETA: 0s - loss: 0.0456 - acc: 0.993 - 0s 236us/sample - loss: 0.0456 - acc: 0.9926\n",
      "Epoch 58/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0130 - acc: 1.000 - ETA: 0s - loss: 0.0181 - acc: 0.996 - ETA: 0s - loss: 0.0268 - acc: 0.996 - ETA: 0s - loss: 0.0306 - acc: 0.997 - ETA: 0s - loss: 0.0361 - acc: 0.995 - ETA: 0s - loss: 0.0372 - acc: 0.993 - ETA: 0s - loss: 0.0429 - acc: 0.992 - 0s 209us/sample - loss: 0.0436 - acc: 0.9926\n",
      "Epoch 59/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0209 - acc: 1.000 - ETA: 0s - loss: 0.0427 - acc: 0.986 - ETA: 0s - loss: 0.0495 - acc: 0.987 - ETA: 0s - loss: 0.0538 - acc: 0.988 - ETA: 0s - loss: 0.0470 - acc: 0.990 - ETA: 0s - loss: 0.0450 - acc: 0.991 - ETA: 0s - loss: 0.0451 - acc: 0.991 - 0s 217us/sample - loss: 0.0457 - acc: 0.9913\n",
      "Epoch 60/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0371 - acc: 1.000 - ETA: 0s - loss: 0.0296 - acc: 1.000 - ETA: 0s - loss: 0.0309 - acc: 0.998 - ETA: 0s - loss: 0.0280 - acc: 0.998 - ETA: 0s - loss: 0.0296 - acc: 0.999 - ETA: 0s - loss: 0.0339 - acc: 0.996 - ETA: 0s - loss: 0.0364 - acc: 0.995 - 0s 221us/sample - loss: 0.0381 - acc: 0.9950\n",
      "Epoch 61/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0242 - acc: 1.000 - ETA: 0s - loss: 0.0516 - acc: 0.990 - ETA: 0s - loss: 0.0370 - acc: 0.994 - ETA: 0s - loss: 0.0366 - acc: 0.995 - ETA: 0s - loss: 0.0392 - acc: 0.994 - ETA: 0s - loss: 0.0378 - acc: 0.995 - ETA: 0s - loss: 0.0408 - acc: 0.994 - 0s 211us/sample - loss: 0.0410 - acc: 0.9938\n",
      "Epoch 62/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0168 - acc: 1.000 - ETA: 0s - loss: 0.0360 - acc: 0.996 - ETA: 0s - loss: 0.0271 - acc: 0.998 - ETA: 0s - loss: 0.0299 - acc: 0.997 - ETA: 0s - loss: 0.0337 - acc: 0.996 - ETA: 0s - loss: 0.0380 - acc: 0.993 - ETA: 0s - loss: 0.0379 - acc: 0.993 - 0s 204us/sample - loss: 0.0377 - acc: 0.9932\n",
      "Epoch 63/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0219 - acc: 1.000 - ETA: 0s - loss: 0.0409 - acc: 0.990 - ETA: 0s - loss: 0.0475 - acc: 0.987 - ETA: 0s - loss: 0.0373 - acc: 0.991 - ETA: 0s - loss: 0.0418 - acc: 0.990 - ETA: 0s - loss: 0.0414 - acc: 0.990 - ETA: 0s - loss: 0.0424 - acc: 0.990 - 0s 212us/sample - loss: 0.0420 - acc: 0.9907\n",
      "Epoch 64/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0044 - acc: 1.000 - ETA: 0s - loss: 0.0544 - acc: 0.987 - ETA: 0s - loss: 0.0450 - acc: 0.989 - ETA: 0s - loss: 0.0365 - acc: 0.991 - ETA: 0s - loss: 0.0445 - acc: 0.989 - ETA: 0s - loss: 0.0422 - acc: 0.990 - ETA: 0s - loss: 0.0412 - acc: 0.990 - 0s 208us/sample - loss: 0.0410 - acc: 0.9901\n",
      "Epoch 65/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1057 - acc: 0.937 - ETA: 0s - loss: 0.0351 - acc: 0.990 - ETA: 0s - loss: 0.0270 - acc: 0.994 - ETA: 0s - loss: 0.0258 - acc: 0.995 - ETA: 0s - loss: 0.0243 - acc: 0.996 - ETA: 0s - loss: 0.0248 - acc: 0.996 - ETA: 0s - loss: 0.0249 - acc: 0.997 - 0s 219us/sample - loss: 0.0248 - acc: 0.9975\n",
      "Epoch 66/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0080 - acc: 1.000 - ETA: 0s - loss: 0.0318 - acc: 0.996 - ETA: 0s - loss: 0.0235 - acc: 0.998 - ETA: 0s - loss: 0.0205 - acc: 0.998 - ETA: 0s - loss: 0.0276 - acc: 0.997 - ETA: 0s - loss: 0.0307 - acc: 0.995 - ETA: 0s - loss: 0.0294 - acc: 0.995 - 0s 213us/sample - loss: 0.0303 - acc: 0.9950\n",
      "Epoch 67/100\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0103 - acc: 1.000 - ETA: 0s - loss: 0.0243 - acc: 0.993 - ETA: 0s - loss: 0.0249 - acc: 0.994 - ETA: 0s - loss: 0.0276 - acc: 0.995 - ETA: 0s - loss: 0.0264 - acc: 0.995 - ETA: 0s - loss: 0.0265 - acc: 0.996 - ETA: 0s - loss: 0.0289 - acc: 0.996 - 0s 213us/sample - loss: 0.0285 - acc: 0.9957\n",
      "Epoch 68/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0053 - acc: 1.000 - ETA: 0s - loss: 0.0210 - acc: 0.996 - ETA: 0s - loss: 0.0207 - acc: 0.996 - ETA: 0s - loss: 0.0230 - acc: 0.996 - ETA: 0s - loss: 0.0193 - acc: 0.996 - ETA: 0s - loss: 0.0228 - acc: 0.996 - ETA: 0s - loss: 0.0241 - acc: 0.996 - 0s 224us/sample - loss: 0.0235 - acc: 0.9969\n",
      "Epoch 69/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0011 - acc: 1.000 - ETA: 0s - loss: 0.0481 - acc: 0.989 - ETA: 0s - loss: 0.0379 - acc: 0.992 - ETA: 0s - loss: 0.0415 - acc: 0.991 - ETA: 0s - loss: 0.0429 - acc: 0.990 - ETA: 0s - loss: 0.0462 - acc: 0.989 - ETA: 0s - loss: 0.0412 - acc: 0.991 - 0s 222us/sample - loss: 0.0404 - acc: 0.9913\n",
      "Epoch 70/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0669 - acc: 1.000 - ETA: 0s - loss: 0.0383 - acc: 0.987 - ETA: 0s - loss: 0.0275 - acc: 0.993 - ETA: 0s - loss: 0.0230 - acc: 0.995 - ETA: 0s - loss: 0.0259 - acc: 0.995 - ETA: 0s - loss: 0.0276 - acc: 0.995 - ETA: 0s - loss: 0.0257 - acc: 0.996 - 0s 234us/sample - loss: 0.0247 - acc: 0.9969\n",
      "Epoch 71/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0038 - acc: 1.000 - ETA: 0s - loss: 0.0438 - acc: 0.988 - ETA: 0s - loss: 0.0297 - acc: 0.994 - ETA: 0s - loss: 0.0288 - acc: 0.994 - ETA: 0s - loss: 0.0284 - acc: 0.994 - ETA: 0s - loss: 0.0343 - acc: 0.992 - ETA: 0s - loss: 0.0337 - acc: 0.992 - 0s 242us/sample - loss: 0.0351 - acc: 0.9913\n",
      "Epoch 72/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0341 - acc: 1.000 - ETA: 0s - loss: 0.0252 - acc: 0.996 - ETA: 0s - loss: 0.0336 - acc: 0.992 - ETA: 0s - loss: 0.0320 - acc: 0.993 - ETA: 0s - loss: 0.0294 - acc: 0.992 - ETA: 0s - loss: 0.0305 - acc: 0.992 - ETA: 0s - loss: 0.0310 - acc: 0.992 - 0s 224us/sample - loss: 0.0297 - acc: 0.9932\n",
      "Epoch 73/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0128 - acc: 1.000 - ETA: 0s - loss: 0.0197 - acc: 0.993 - ETA: 0s - loss: 0.0164 - acc: 0.996 - ETA: 0s - loss: 0.0231 - acc: 0.995 - ETA: 0s - loss: 0.0237 - acc: 0.996 - ETA: 0s - loss: 0.0228 - acc: 0.996 - ETA: 0s - loss: 0.0249 - acc: 0.996 - 0s 221us/sample - loss: 0.0273 - acc: 0.9963\n",
      "Epoch 74/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0033 - acc: 1.000 - ETA: 0s - loss: 0.0194 - acc: 0.993 - ETA: 0s - loss: 0.0198 - acc: 0.994 - ETA: 0s - loss: 0.0203 - acc: 0.995 - ETA: 0s - loss: 0.0201 - acc: 0.995 - ETA: 0s - loss: 0.0209 - acc: 0.995 - ETA: 0s - loss: 0.0210 - acc: 0.995 - 0s 229us/sample - loss: 0.0224 - acc: 0.9950\n",
      "Epoch 75/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0047 - acc: 1.000 - ETA: 0s - loss: 0.0361 - acc: 0.993 - ETA: 0s - loss: 0.0295 - acc: 0.994 - ETA: 0s - loss: 0.0275 - acc: 0.995 - ETA: 0s - loss: 0.0247 - acc: 0.996 - ETA: 0s - loss: 0.0296 - acc: 0.994 - ETA: 0s - loss: 0.0293 - acc: 0.994 - 0s 222us/sample - loss: 0.0293 - acc: 0.9944\n",
      "Epoch 76/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0087 - acc: 1.000 - ETA: 0s - loss: 0.0307 - acc: 0.993 - ETA: 0s - loss: 0.0236 - acc: 0.994 - ETA: 0s - loss: 0.0218 - acc: 0.995 - ETA: 0s - loss: 0.0217 - acc: 0.997 - ETA: 0s - loss: 0.0235 - acc: 0.996 - ETA: 0s - loss: 0.0252 - acc: 0.994 - 0s 228us/sample - loss: 0.0261 - acc: 0.9944\n",
      "Epoch 77/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0015 - acc: 1.000 - ETA: 0s - loss: 0.0070 - acc: 1.000 - ETA: 0s - loss: 0.0162 - acc: 0.996 - ETA: 0s - loss: 0.0226 - acc: 0.994 - ETA: 0s - loss: 0.0225 - acc: 0.994 - ETA: 0s - loss: 0.0252 - acc: 0.994 - ETA: 0s - loss: 0.0276 - acc: 0.993 - 0s 224us/sample - loss: 0.0259 - acc: 0.9944\n",
      "Epoch 78/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0265 - acc: 1.000 - ETA: 0s - loss: 0.0119 - acc: 1.000 - ETA: 0s - loss: 0.0151 - acc: 0.998 - ETA: 0s - loss: 0.0228 - acc: 0.995 - ETA: 0s - loss: 0.0220 - acc: 0.996 - ETA: 0s - loss: 0.0262 - acc: 0.994 - ETA: 0s - loss: 0.0246 - acc: 0.995 - 0s 210us/sample - loss: 0.0245 - acc: 0.9957\n",
      "Epoch 79/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0023 - acc: 1.000 - ETA: 0s - loss: 0.0348 - acc: 0.986 - ETA: 0s - loss: 0.0248 - acc: 0.992 - ETA: 0s - loss: 0.0345 - acc: 0.990 - ETA: 0s - loss: 0.0337 - acc: 0.991 - ETA: 0s - loss: 0.0348 - acc: 0.990 - ETA: 0s - loss: 0.0325 - acc: 0.991 - 0s 218us/sample - loss: 0.0340 - acc: 0.9907\n",
      "Epoch 80/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0124 - acc: 1.000 - ETA: 0s - loss: 0.0216 - acc: 0.996 - ETA: 0s - loss: 0.0177 - acc: 0.998 - ETA: 0s - loss: 0.0147 - acc: 0.998 - ETA: 0s - loss: 0.0243 - acc: 0.997 - ETA: 0s - loss: 0.0266 - acc: 0.997 - ETA: 0s - loss: 0.0271 - acc: 0.997 - 0s 229us/sample - loss: 0.0283 - acc: 0.9963\n",
      "Epoch 81/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1378 - acc: 0.968 - ETA: 0s - loss: 0.0225 - acc: 0.996 - ETA: 0s - loss: 0.0216 - acc: 0.996 - ETA: 0s - loss: 0.0211 - acc: 0.995 - ETA: 0s - loss: 0.0241 - acc: 0.995 - ETA: 0s - loss: 0.0262 - acc: 0.994 - 0s 204us/sample - loss: 0.0270 - acc: 0.9938\n",
      "Epoch 82/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0022 - acc: 1.000 - ETA: 0s - loss: 0.0233 - acc: 0.993 - ETA: 0s - loss: 0.0198 - acc: 0.995 - ETA: 0s - loss: 0.0206 - acc: 0.994 - ETA: 0s - loss: 0.0235 - acc: 0.994 - ETA: 0s - loss: 0.0263 - acc: 0.994 - ETA: 0s - loss: 0.0253 - acc: 0.994 - 0s 216us/sample - loss: 0.0253 - acc: 0.9950\n",
      "Epoch 83/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0020 - acc: 1.000 - ETA: 0s - loss: 0.0106 - acc: 1.000 - ETA: 0s - loss: 0.0272 - acc: 0.992 - ETA: 0s - loss: 0.0245 - acc: 0.993 - ETA: 0s - loss: 0.0266 - acc: 0.992 - ETA: 0s - loss: 0.0265 - acc: 0.993 - ETA: 0s - loss: 0.0263 - acc: 0.993 - 0s 210us/sample - loss: 0.0256 - acc: 0.9932\n",
      "Epoch 84/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0052 - acc: 1.000 - ETA: 0s - loss: 0.0170 - acc: 0.993 - ETA: 0s - loss: 0.0239 - acc: 0.995 - ETA: 0s - loss: 0.0213 - acc: 0.995 - ETA: 0s - loss: 0.0208 - acc: 0.996 - ETA: 0s - loss: 0.0247 - acc: 0.995 - ETA: 0s - loss: 0.0257 - acc: 0.994 - 0s 218us/sample - loss: 0.0245 - acc: 0.9950\n",
      "Epoch 85/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0108 - acc: 1.000 - ETA: 0s - loss: 0.0206 - acc: 0.993 - ETA: 0s - loss: 0.0170 - acc: 0.995 - ETA: 0s - loss: 0.0163 - acc: 0.996 - ETA: 0s - loss: 0.0197 - acc: 0.996 - ETA: 0s - loss: 0.0233 - acc: 0.995 - ETA: 0s - loss: 0.0226 - acc: 0.995 - 0s 207us/sample - loss: 0.0224 - acc: 0.9957\n",
      "Epoch 86/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0127 - acc: 1.000 - ETA: 0s - loss: 0.0198 - acc: 0.993 - ETA: 0s - loss: 0.0257 - acc: 0.994 - ETA: 0s - loss: 0.0239 - acc: 0.995 - ETA: 0s - loss: 0.0234 - acc: 0.996 - ETA: 0s - loss: 0.0210 - acc: 0.997 - 0s 202us/sample - loss: 0.0244 - acc: 0.9957\n",
      "Epoch 87/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0338 - acc: 1.000 - ETA: 0s - loss: 0.0098 - acc: 1.000 - ETA: 0s - loss: 0.0114 - acc: 0.998 - ETA: 0s - loss: 0.0123 - acc: 0.997 - ETA: 0s - loss: 0.0181 - acc: 0.997 - ETA: 0s - loss: 0.0185 - acc: 0.997 - ETA: 0s - loss: 0.0184 - acc: 0.996 - 0s 205us/sample - loss: 0.0183 - acc: 0.9969\n",
      "Epoch 88/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0060 - acc: 1.000 - ETA: 0s - loss: 0.0243 - acc: 0.996 - ETA: 0s - loss: 0.0217 - acc: 0.998 - ETA: 0s - loss: 0.0187 - acc: 0.997 - ETA: 0s - loss: 0.0220 - acc: 0.996 - ETA: 0s - loss: 0.0206 - acc: 0.996 - 0s 206us/sample - loss: 0.0212 - acc: 0.9963\n",
      "Epoch 89/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0039 - acc: 1.000 - ETA: 0s - loss: 0.0148 - acc: 0.996 - ETA: 0s - loss: 0.0134 - acc: 0.996 - ETA: 0s - loss: 0.0140 - acc: 0.996 - ETA: 0s - loss: 0.0179 - acc: 0.995 - ETA: 0s - loss: 0.0197 - acc: 0.995 - ETA: 0s - loss: 0.0195 - acc: 0.996 - 0s 215us/sample - loss: 0.0214 - acc: 0.9950\n",
      "Epoch 90/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0054 - acc: 1.000 - ETA: 0s - loss: 0.0116 - acc: 0.996 - ETA: 0s - loss: 0.0146 - acc: 0.994 - ETA: 0s - loss: 0.0173 - acc: 0.996 - ETA: 0s - loss: 0.0220 - acc: 0.996 - ETA: 0s - loss: 0.0230 - acc: 0.995 - ETA: 0s - loss: 0.0223 - acc: 0.996 - 0s 241us/sample - loss: 0.0223 - acc: 0.9957\n",
      "Epoch 91/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1056 - acc: 0.937 - ETA: 0s - loss: 0.0407 - acc: 0.984 - ETA: 0s - loss: 0.0328 - acc: 0.989 - ETA: 0s - loss: 0.0293 - acc: 0.991 - ETA: 0s - loss: 0.0259 - acc: 0.992 - ETA: 0s - loss: 0.0298 - acc: 0.991 - ETA: 0s - loss: 0.0273 - acc: 0.992 - 0s 228us/sample - loss: 0.0273 - acc: 0.9920\n",
      "Epoch 92/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0247 - acc: 1.000 - ETA: 0s - loss: 0.0153 - acc: 1.000 - ETA: 0s - loss: 0.0143 - acc: 0.998 - ETA: 0s - loss: 0.0185 - acc: 0.996 - ETA: 0s - loss: 0.0215 - acc: 0.994 - ETA: 0s - loss: 0.0193 - acc: 0.995 - ETA: 0s - loss: 0.0196 - acc: 0.994 - 0s 225us/sample - loss: 0.0184 - acc: 0.9950\n",
      "Epoch 93/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.1112 - acc: 0.968 - ETA: 0s - loss: 0.0164 - acc: 0.996 - ETA: 0s - loss: 0.0198 - acc: 0.996 - ETA: 0s - loss: 0.0227 - acc: 0.994 - ETA: 0s - loss: 0.0204 - acc: 0.996 - ETA: 0s - loss: 0.0223 - acc: 0.995 - ETA: 0s - loss: 0.0217 - acc: 0.995 - 0s 224us/sample - loss: 0.0211 - acc: 0.9957\n",
      "Epoch 94/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0137 - acc: 1.000 - ETA: 0s - loss: 0.0294 - acc: 0.987 - ETA: 0s - loss: 0.0309 - acc: 0.989 - ETA: 0s - loss: 0.0284 - acc: 0.992 - ETA: 0s - loss: 0.0318 - acc: 0.990 - ETA: 0s - loss: 0.0296 - acc: 0.991 - ETA: 0s - loss: 0.0271 - acc: 0.992 - ETA: 0s - loss: 0.0273 - acc: 0.993 - 0s 250us/sample - loss: 0.0271 - acc: 0.9926\n",
      "Epoch 95/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0290 - acc: 1.000 - ETA: 0s - loss: 0.0313 - acc: 0.993 - ETA: 0s - loss: 0.0259 - acc: 0.994 - ETA: 0s - loss: 0.0249 - acc: 0.993 - ETA: 0s - loss: 0.0231 - acc: 0.993 - ETA: 0s - loss: 0.0223 - acc: 0.993 - ETA: 0s - loss: 0.0211 - acc: 0.994 - 0s 228us/sample - loss: 0.0206 - acc: 0.9944\n",
      "Epoch 96/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0034 - acc: 1.000 - ETA: 0s - loss: 0.0181 - acc: 0.996 - ETA: 0s - loss: 0.0183 - acc: 0.996 - ETA: 0s - loss: 0.0224 - acc: 0.995 - ETA: 0s - loss: 0.0187 - acc: 0.996 - ETA: 0s - loss: 0.0172 - acc: 0.996 - ETA: 0s - loss: 0.0171 - acc: 0.997 - 0s 217us/sample - loss: 0.0185 - acc: 0.9969\n",
      "Epoch 97/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0029 - acc: 1.000 - ETA: 0s - loss: 0.0310 - acc: 0.993 - ETA: 0s - loss: 0.0245 - acc: 0.994 - ETA: 0s - loss: 0.0231 - acc: 0.994 - ETA: 0s - loss: 0.0245 - acc: 0.994 - ETA: 0s - loss: 0.0252 - acc: 0.993 - ETA: 0s - loss: 0.0252 - acc: 0.992 - 0s 222us/sample - loss: 0.0256 - acc: 0.9932\n",
      "Epoch 98/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0577 - acc: 1.000 - ETA: 0s - loss: 0.0258 - acc: 0.993 - ETA: 0s - loss: 0.0197 - acc: 0.994 - ETA: 0s - loss: 0.0181 - acc: 0.996 - ETA: 0s - loss: 0.0186 - acc: 0.996 - ETA: 0s - loss: 0.0185 - acc: 0.996 - ETA: 0s - loss: 0.0190 - acc: 0.996 - 0s 219us/sample - loss: 0.0196 - acc: 0.9957\n",
      "Epoch 99/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0024 - acc: 1.000 - ETA: 0s - loss: 0.0053 - acc: 1.000 - ETA: 0s - loss: 0.0039 - acc: 1.000 - ETA: 0s - loss: 0.0094 - acc: 0.998 - ETA: 0s - loss: 0.0111 - acc: 0.999 - ETA: 0s - loss: 0.0135 - acc: 0.998 - ETA: 0s - loss: 0.0152 - acc: 0.997 - 0s 239us/sample - loss: 0.0188 - acc: 0.9963\n",
      "Epoch 100/100\n",
      "1616/1616 [==============================] - ETA: 0s - loss: 0.0017 - acc: 1.000 - ETA: 0s - loss: 0.0042 - acc: 1.000 - ETA: 0s - loss: 0.0188 - acc: 0.994 - ETA: 0s - loss: 0.0242 - acc: 0.992 - ETA: 0s - loss: 0.0292 - acc: 0.991 - ETA: 0s - loss: 0.0324 - acc: 0.991 - ETA: 0s - loss: 0.0301 - acc: 0.993 - 0s 220us/sample - loss: 0.0295 - acc: 0.9932\n"
     ]
    }
   ],
   "source": [
    "train_core()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### optional chat test\n",
    "chat(movichtr_model_path)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Actions\n",
    "\n",
    "Generally, it is useful for a conversational system to do more than just answer according to a set of templates. Most of the times users use chatbots to execute specific tasks; for that there are **actions**. Creating custom actions allow you to answer a user's intent in better ways than just replying something (e.g.: you may add an event to a calendar or check the weather).\n",
    "\n",
    "When there is the need to execute a custom action, Rasa calls an endpoint where a server should be up and ready to answer accordingly (Rasa provides a python SDK to help with that part). To inform the system where the server is running, a file with the endpoint of the server should be created.\n",
    "\n",
    "To know more about actions check this link: https://rasa.com/docs/rasa/core/actions/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Only run it to overwrite the endpoints if not set up previously\n",
    "endpoints= \"\"\"\n",
    "action_endpoint:\n",
    "    url: http://localhost:{}/webhook\n",
    "\"\"\".format(PORT)\n",
    "\n",
    "write_to_file(movichtr_endpoints_file, endpoints)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "api_handler = \"\"\"\n",
    "import requests\n",
    "import json\n",
    "\n",
    "#Global Variables\n",
    "API_KEY = 'HIDDEN'\n",
    "BASE_URL = 'https://api.themoviedb.org/3/'\n",
    "\n",
    "#defining subject entity search queries\n",
    "\n",
    "#popular movies\n",
    "def get_popularity_by_year(year):\n",
    "    #adding API endpoint (discover)\n",
    "    url = BASE_URL + 'discover/movie'\n",
    "    \n",
    "    #parameter specification\n",
    "    params = {'api_key': API_KEY, 'sort_by' : 'popularity.desc', 'include_adult' : 'false', 'primary_release_year' : year}\n",
    "    \n",
    "    #perform request\n",
    "    response = requests.get(url, params = params).text\n",
    "    \n",
    "    response_json = json.loads(response)\n",
    "    \n",
    "    #get results array\n",
    "    results = response_json['results']\n",
    "    \n",
    "    return results\n",
    "    \n",
    "#get known movies by crew person (example: actor/ director)\n",
    "def get_movies_by_crew(query):\n",
    "    url = BASE_URL + 'search/multi'\n",
    "    \n",
    "    params = {'api_key': API_KEY, 'query' : query, 'include_adult' : 'false'}  \n",
    "    \n",
    "    response = requests.get(url, params = params).text\n",
    "    \n",
    "    response_json = json.loads(response)\n",
    "    \n",
    "    results = response_json['results']\n",
    "    \n",
    "    movies = parseMoviesWith(results, query)\n",
    "    \n",
    "    return movies\n",
    "  \n",
    "\n",
    "#get credits by movie title (example: who directed/acted in movie?)\n",
    "def get_crew_by_movie(query):\n",
    "    url = BASE_URL + 'movie/{}/credits'.format(query)\n",
    "    \n",
    "    params = {'api_key': API_KEY}  \n",
    "    \n",
    "    response = requests.get(url, params = params).text\n",
    "    \n",
    "    response_json = json.loads(response)\n",
    "    \n",
    "    crew = response_json['crew']\n",
    "    \n",
    "    return crew   \n",
    "    \n",
    "#get cast by movie title (example: who directed/acted in movie?)\n",
    "def get_cast_by_movie(query):\n",
    "    url = BASE_URL + 'movie/{}/credits'.format(query)\n",
    "    \n",
    "    params = {'api_key': API_KEY}  \n",
    "    \n",
    "    response = requests.get(url, params = params).text\n",
    "    \n",
    "    response_json = json.loads(response)\n",
    "    \n",
    "    cast = response_json['cast']\n",
    "    \n",
    "    return cast  \n",
    "    \n",
    "\n",
    "#get movie attributes by movie_title (example: rating/release_year)\n",
    "def get_attributes_by_movie_title(query):\n",
    "    url = BASE_URL + 'search/movie'\n",
    "    \n",
    "    params = {'api_key': API_KEY, 'query' : query, 'include_adult' : 'false'}\n",
    "    \n",
    "    response = requests.get(url, params = params).text\n",
    "    \n",
    "    response_json = json.loads(response)\n",
    "    \n",
    "    if (response_json.get('results') == None):\n",
    "        print(\"Error requesting attributes for query\")\n",
    "        return None\n",
    "    \n",
    "    results = response_json['results']\n",
    "    \n",
    "    return results\n",
    "\n",
    "\n",
    "#finds known for movies for given person\n",
    "def parseMoviesWith(json_data, person):\n",
    "    movies = []\n",
    "\n",
    "    for results in json_data:\n",
    "        if results['media_type'] == 'person':\n",
    "            known_for = results['known_for']\n",
    "            for movie in known_for:\n",
    "                title = movie['original_title']\n",
    "                movies.append(title)\n",
    "\n",
    "    return movies\n",
    "    \n",
    "#finds movies similar to inputted title\n",
    "def get_similar_movies(query):\n",
    "    url = BASE_URL + 'movie/{}/similar'.format(query)\n",
    "    \n",
    "    params = {'api_key': API_KEY}  \n",
    "    \n",
    "    response = requests.get(url, params = params).text\n",
    "    \n",
    "    response_json = json.loads(response)\n",
    "    \n",
    "    if (response_json.get('results') == None):\n",
    "        print(\"Error requesting attributes for query\")\n",
    "        return None\n",
    "        \n",
    "    results = response_json['results']\n",
    "    \n",
    "    return results\n",
    "    \n",
    "\"\"\"\n",
    "\n",
    "write_to_file(movichtr_api_file, api_handler)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#show files in the project directory, optional\n",
    "print(os.listdir(\"./\" + movichtr_project))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "movichtr_action = \"\"\"\n",
    "from rasa_core_sdk import Action\n",
    "import requests\n",
    "import json\n",
    "import api_handler as mapi\n",
    "import logging\n",
    "logger = logging.getLogger(__name__)\n",
    "\n",
    "#searching movies by actor\n",
    "class ActionSearchMoviesByActor(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_movie_search_by_actor'\n",
    "\n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "        actor = tracker.get_slot('actor')\n",
    "        \n",
    "        if (actor is None or len(actor) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what movie are you talking about. Please try again\")\n",
    "            return\n",
    "            \n",
    "        movies = mapi.get_movies_by_crew(actor)\n",
    "        if len(movies) == 0:\n",
    "            dispatcher.utter_message(\"No movies with {} were found\".format(actor))\n",
    "        else: \n",
    "            dispatcher.utter_message(\"{} is known for the following movies:\".format(actor))\n",
    "            i = 1\n",
    "            for movie in movies:\n",
    "                dispatcher.utter_message(str(i) + \". \" + movie)\n",
    "                i = i+1\n",
    "\n",
    "        return[]\n",
    "        \n",
    "#______________________________________________________________________        \n",
    "#searching movies by director\n",
    "class ActionSearchMoviesByDirector(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_movie_search_by_director'\n",
    "\n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "    \n",
    "        director = tracker.get_slot('director')\n",
    "        \n",
    "        if (director is None or len(director) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what director are you talking about. Please try again\")\n",
    "            return\n",
    "            \n",
    "        movies = mapi.get_movies_by_crew(director)\n",
    "        if len(movies) == 0:\n",
    "            dispatcher.utter_message(\"No movies with {} were found\".format(director))\n",
    "        else: \n",
    "            dispatcher.utter_message(\"{} is known for the following movies:\".format(director))\n",
    "            i = 1\n",
    "            for movie in movies:\n",
    "                dispatcher.utter_message(str(i) + \". \" + movie)\n",
    "                i = i+1\n",
    "\n",
    "        return[]\n",
    "        \n",
    "#______________________________________________________________________           \n",
    "#searching movies by release\n",
    "class ActionSearchReleaseYearForMovie(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_search_release_year_for_movie'\n",
    "\n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "        movie_title = tracker.get_slot('movie_title')\n",
    "        \n",
    "        if (movie_title is None or len(movie_title) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what movie are you talking about. Please try again\")\n",
    "            return\n",
    "            \n",
    "        results = mapi.get_attributes_by_movie_title(movie_title)\n",
    "        if len(results) == 0:\n",
    "            dispatcher.utter_message(\"I can't find anything with title {}\".format(movie_title))\n",
    "        else:\n",
    "            movie = results[0]\n",
    "            original_title = movie['original_title']\n",
    "            release_year = movie['release_date'].split('-')[0]\n",
    "            \n",
    "            dispatcher.utter_message(\"{title} was released in {year}\".format(title=original_title, year=release_year))\n",
    "            \n",
    "        return[]\n",
    "      \n",
    "#______________________________________________________________________         \n",
    "#searching for movie plot/overview       \n",
    "class ActionSearchOverviewForMovie(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_search_overview_for_movie'\n",
    "        \n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "        movie_title = tracker.get_slot('movie_title')\n",
    "        \n",
    "        if (movie_title is None or len(movie_title) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what movie are you talking about. Please try again\")\n",
    "            return\n",
    "        \n",
    "        results = mapi.get_attributes_by_movie_title(movie_title)\n",
    "        \n",
    "        if len(results) == 0:\n",
    "            dispatcher.utter_message(\"I can't find anything with title {}\".format(movie_title))\n",
    "        else:\n",
    "            movie = results[0]\n",
    "            original_title = movie['original_title']\n",
    "            overview = movie['overview']\n",
    "            \n",
    "            dispatcher.utter_message(\"{title} has the following overview {overview}\".format(title=original_title, overview=overview))\n",
    "            \n",
    "        return[]\n",
    "\n",
    "#______________________________________________________________________   \n",
    "#searching for movie ratings by movie title\n",
    "class ActionSearchRatingForMovie(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_search_rating_for_movie'\n",
    "        \n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "    \n",
    "        movie_title = tracker.get_slot('movie_title')\n",
    "        \n",
    "        if (movie_title is None or len(movie_title) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what movie are you talking about. Please try again\")\n",
    "            return\n",
    "        \n",
    "        results = mapi.get_attributes_by_movie_title(movie_title)\n",
    "\n",
    "        if len(results) == 0:\n",
    "            dispatcher.utter_message(\"I can't find anything with title {}\".format(movie_title))\n",
    "        else:\n",
    "            movie = results[0]\n",
    "            original_title = movie['original_title']\n",
    "            rating = movie['vote_average']\n",
    "            \n",
    "            dispatcher.utter_message(\"{title} has the following rating {rating}\".format(title=original_title, rating=rating))\n",
    "            \n",
    "        return[]\n",
    "\n",
    "#______________________________________________________________________   \n",
    "#search popular movies by year\n",
    "class ActionSearchPopularityByYear(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_search_popularity_by_year'\n",
    "        \n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "        year = tracker.get_slot('release')\n",
    "        \n",
    "        if (year is None or len(year) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what date are you talking about. Please try again\")\n",
    "            return\n",
    "            \n",
    "        movies = mapi.get_popularity_by_year(year)\n",
    "        \n",
    "        dispatcher.utter_message(\"These are the most popular movies of {}\".format(year))\n",
    "        \n",
    "        i = 1\n",
    "        for movie in movies:\n",
    "            dispatcher.utter_message(str(i) + \". \" + movie['original_title'] + \" with rating \" + str(movie['vote_average']))\n",
    "            i = i + 1\n",
    "            \n",
    "        return[]\n",
    "        \n",
    "#______________________________________________________________________   \n",
    "#search director for specific movie\n",
    "class ActionSearchMovieDirector(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_search_movie_director'\n",
    "        \n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "        movie_title = tracker.get_slot('movie_title')\n",
    "        \n",
    "        if (movie_title is None or len(movie_title) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what movie are you talking about. Please try again\")\n",
    "            return \n",
    "        \n",
    "        results = mapi.get_attributes_by_movie_title(movie_title)\n",
    "        crew = mapi.get_crew_by_movie(results[0]['id'])\n",
    "        \n",
    "        if len(results) == 0:\n",
    "            dispatcher.utter_message(\"I can't find anything with title {}\".format(movie_title))\n",
    "        else:\n",
    "            for member in crew:\n",
    "                if (member['job'] == 'Director'):\n",
    "                    dispatcher.utter_message(results[0]['original_title'] + ' was directed by ' + member['name'])\n",
    "                    \n",
    "        return[]\n",
    "#______________________________________________________________________   \n",
    "#search actors for specific movie\n",
    "class ActionSearchMovieActors(Action):\n",
    "\n",
    "    def name(self):\n",
    "        return 'action_search_movie_actors'\n",
    "        \n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "        movie_title = tracker.get_slot('movie_title')\n",
    "        \n",
    "        if (movie_title is None or len(movie_title) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what movie are you talking about. Please try again\")\n",
    "            return\n",
    "            \n",
    "        results = mapi.get_attributes_by_movie_title(movie_title)\n",
    "        \n",
    "        cast = mapi.get_cast_by_movie(results[0]['id'])\n",
    "        \n",
    "        if len(results) == 0:\n",
    "            dispatcher.utter_message(\"I can't find anything with title {}\".format(movie_title))\n",
    "        else:\n",
    "            dispatcher.utter_message(\"These actors starred in {}\".format(movie_title))\n",
    "            i = 1\n",
    "            for member in cast:\n",
    "                actor = member['name']\n",
    "                character = member['character']\n",
    "                dispatcher.utter_message(str(i) + \". \" + actor + ' starred with character ' + character)\n",
    "                i = i + 1   \n",
    "                if (i > 10):\n",
    "                    break\n",
    "                    \n",
    "        return[]\n",
    "#______________________________________________________________________   \n",
    "#search similar movies\n",
    "class ActionSearchSimilarMovies(Action):\n",
    "    \n",
    "    def name(self):\n",
    "        return 'action_search_similar_movies'\n",
    "        \n",
    "    def run(self, dispatcher, tracker, domain):\n",
    "        movie_title = tracker.get_slot('movie_title')\n",
    "        \n",
    "        if (movie_title is None or len(movie_title) == 0):\n",
    "            dispatcher.utter_message(\"Sorry, I didn't get what movie are you talking about. Please try again\")\n",
    "            return\n",
    "            \n",
    "        results = mapi.get_attributes_by_movie_title(movie_title)\n",
    "        \n",
    "        similar = mapi.get_similar_movies(results[0]['id'])\n",
    "        \n",
    "        if len(results) == 0:\n",
    "            dispatcher.utter_message(\"I can't find anything similar to title {}\".format(movie_title))\n",
    "        else:\n",
    "            dispatcher.utter_message(\"These are similar to {}\".format(movie_title))\n",
    "            i = 1\n",
    "            for member in similar:\n",
    "                suggestion = member['original_title']\n",
    "                release = member['release_date']\n",
    "                dispatcher.utter_message(str(i) + \". \" + suggestion + ' released in  ' + release)\n",
    "                i = i + 1   \n",
    "                if (i > 10):\n",
    "                    break\n",
    "                    \n",
    "        return[]\n",
    "        \n",
    "                                   \n",
    "\"\"\"\n",
    "write_to_file(movichtr_actions_file, movichtr_action)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### To take advantage of the newly created action, we have to start the rasa_sdk server. To do that, open a Jupyter terminal, go to the example project folder and run the command:\n",
    "\n",
    "rasa run actions\n",
    "\n",
    "### and finally on a third\n",
    "rasa x\n",
    "\n",
    "## Because I often forget and skip this step another reminder: RASA RUN ACTIONS\n",
    "\n",
    "### To run the assistant from here just keep the slack related code commented out"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "def load_assistant():\n",
    "    messages = [\"Hi! you can chat in this window. Type 'stop' to end the conversation.\"]\n",
    "    interpreter = NaturalLanguageInterpreter.create(movichtr_models_output + movichtr_model_name + \"/nlu/\")\n",
    "    endpoint = EndpointConfig('http://localhost:{}/webhook'.format(PORT))\n",
    "    agent = Agent.load(movichtr_model_path, interpreter=interpreter, action_endpoint = endpoint)\n",
    "    \n",
    "    \n",
    "    input_channel = SlackInput(\n",
    "    slack_token=\"xoxb-720064421394-807667952307-WjHRR4bMmBeTtpYD2QlQIQT9\",\n",
    "    slack_channel=\"#challenge-bot-testing\")\n",
    "    s = agent.handle_channels([input_channel], 5004)\n",
    "    \n",
    "    \n",
    "    print(\"Your bot is ready to talk! Type your messages here or send 'stop'\")\n",
    "    while True:\n",
    "        user_message = input()\n",
    "        if user_message == 'stop':\n",
    "            break\n",
    "            \n",
    "        loop = asyncio.get_event_loop()\n",
    "        responses = loop.run_until_complete(agent.handle_text(user_message))\n",
    "        for response in responses:\n",
    "            for response_type, value in response.items():\n",
    "                if response_type == \"text\":\n",
    "                    print(value)\n",
    "\n",
    "                if response_type == \"image\":\n",
    "                    image = Image(url=value)\n",
    "                    display(image)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Testing the assistant"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "load_assistant()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}