Seq2seq_tf_version_1_3.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "**The code was inspired from the official documnetation of TensorFlow**\n"
      ],
      "metadata": {
        "id": "p0l9gL_3l9do"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JWM5pH9_l6pG"
      },
      "outputs": [],
      "source": [
        "#------------------------------------------#\n",
        "#     Projet Machine Learning Avancé       #\n",
        "#------------------------------------------#\n",
        "#               Master 2 ISI               #\n",
        "\n",
        "#     Membres du Groupe :                  #\n",
        "#     - Tarek BENAISSA                     #\n",
        "#     - Mhamed BENABID                     #\n",
        "#     - Zahra Hafida BENSLIMANE            #\n",
        "#     - Mohammed El Amine BENBRAHIM         #\n",
        "#------------------------------------------#"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uTgDiD57HTe6",
        "outputId": "64bda0bd-c52e-4b6e-9ae7-7c405f8628b7",
        "scrolled": true
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: tensorflow_text in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (2.9.0)\n",
            "Requirement already satisfied: tensorflow<2.10,>=2.9.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow_text) (2.9.2)\n",
            "Requirement already satisfied: tensorflow-hub>=0.8.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow_text) (0.12.0)\n",
            "Requirement already satisfied: google-pasta>=0.1.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (0.2.0)\n",
            "Requirement already satisfied: keras-preprocessing>=1.1.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.1.2)\n",
            "Requirement already satisfied: numpy>=1.20 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.23.1)\n",
            "Requirement already satisfied: typing-extensions>=3.6.6 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (4.3.0)\n",
            "Requirement already satisfied: astunparse>=1.6.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.6.3)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.1.0)\n",
            "Requirement already satisfied: six>=1.12.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.16.0)\n",
            "Requirement already satisfied: wrapt>=1.11.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.14.1)\n",
            "Requirement already satisfied: keras<2.10.0,>=2.9.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (2.9.0)\n",
            "Requirement already satisfied: setuptools in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (63.4.1)\n",
            "Requirement already satisfied: opt-einsum>=2.3.2 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (3.3.0)\n",
            "Requirement already satisfied: tensorflow-estimator<2.10.0,>=2.9.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (2.9.0)\n",
            "Requirement already satisfied: gast<=0.4.0,>=0.2.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (0.4.0)\n",
            "Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (0.26.0)\n",
            "Requirement already satisfied: protobuf<3.20,>=3.9.2 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (3.19.1)\n",
            "Requirement already satisfied: tensorboard<2.10,>=2.9 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (2.9.1)\n",
            "Requirement already satisfied: packaging in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (21.3)\n",
            "Requirement already satisfied: h5py>=2.9.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (3.6.0)\n",
            "Requirement already satisfied: grpcio<2.0,>=1.24.3 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.42.0)\n",
            "Requirement already satisfied: flatbuffers>=2.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (2.0)\n",
            "Requirement already satisfied: absl-py>=1.0.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorflow<2.10,>=2.9.0->tensorflow_text) (1.0.0)\n",
            "Requirement already satisfied: wheel<1.0,>=0.23.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from astunparse>=1.6.0->tensorflow<2.10,>=2.9.0->tensorflow_text) (0.37.1)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (3.3.4)\n",
            "Requirement already satisfied: werkzeug>=1.0.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (2.0.3)\n",
            "Requirement already satisfied: google-auth<3,>=1.6.3 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (1.33.0)\n",
            "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (0.4.4)\n",
            "Requirement already satisfied: requests<3,>=2.21.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (2.28.1)\n",
            "Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (0.6.1)\n",
            "Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (1.6.0)\n",
            "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from packaging->tensorflow<2.10,>=2.9.0->tensorflow_text) (3.0.9)\n",
            "Requirement already satisfied: rsa<5,>=3.1.4 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (4.7.2)\n",
            "Requirement already satisfied: cachetools<5.0,>=2.0.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (4.2.2)\n",
            "Requirement already satisfied: pyasn1-modules>=0.2.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from google-auth<3,>=1.6.3->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (0.2.8)\n",
            "Requirement already satisfied: requests-oauthlib>=0.7.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (1.3.0)\n",
            "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from requests<3,>=2.21.0->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (1.26.11)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from requests<3,>=2.21.0->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (2022.9.24)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from requests<3,>=2.21.0->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (3.3)\n",
            "Requirement already satisfied: charset-normalizer<3,>=2 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from requests<3,>=2.21.0->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (2.0.4)\n",
            "Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (0.4.8)\n",
            "Requirement already satisfied: oauthlib>=3.0.0 in /opt/conda/envs/Python-3.10/lib/python3.10/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard<2.10,>=2.9->tensorflow<2.10,>=2.9.0->tensorflow_text) (3.2.1)\n"
          ]
        }
      ],
      "source": [
        "!pip install tensorflow_text"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "R2Lr5TLIHeKq"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "\n",
        "import typing\n",
        "from typing import Any, Tuple\n",
        "\n",
        "import tensorflow as tf\n",
        "\n",
        "import tensorflow_text as tf_text\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import matplotlib.ticker as ticker"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "--QlcPNMHnQR"
      },
      "outputs": [],
      "source": [
        "# cette variable est utilisée pour basculer entre les implémentations personnalisées ( qu'on crée à partir de zéro) et intégrées.\n",
        "use_builtins = True"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "t0uzprOml6pN"
      },
      "outputs": [],
      "source": [
        "class ShapeChecker():\n",
        "    def __init__(self):\n",
        "        # Keep a cache of every axis-name seen\n",
        "        self.shapes = {}\n",
        "\n",
        "    def __call__(self, tensor, names, broadcast=False):\n",
        "        if not tf.executing_eagerly():\n",
        "            return\n",
        "\n",
        "        if isinstance(names, str):\n",
        "            names = (names,)\n",
        "\n",
        "        shape = tf.shape(tensor)\n",
        "        rank = tf.rank(tensor)\n",
        "\n",
        "        if rank != len(names):\n",
        "            raise ValueError(f'Rank mismatch:\\n'\n",
        "                           f'    found {rank}: {shape.numpy()}\\n'\n",
        "                           f'    expected {len(names)}: {names}\\n')\n",
        "\n",
        "        for i, name in enumerate(names):\n",
        "            if isinstance(name, int):\n",
        "                old_dim = name\n",
        "            else:\n",
        "                old_dim = self.shapes.get(name, None)\n",
        "            new_dim = shape[i]\n",
        "\n",
        "            if (broadcast and new_dim == 1):\n",
        "                continue\n",
        "\n",
        "            if old_dim is None:\n",
        "                # If the axis name is new, add its length to the cache.\n",
        "                self.shapes[name] = new_dim\n",
        "                continue\n",
        "\n",
        "            if new_dim != old_dim:\n",
        "                raise ValueError(f\"Shape mismatch for dimension: '{name}'\\n\"\n",
        "                             f\"    found: {new_dim}\\n\"\n",
        "                             f\"    expected: {old_dim}\\n\")\n",
        "                             \n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "SRLltQepHoZU"
      },
      "outputs": [],
      "source": [
        "# format des données en entrée\n",
        "# input target\n",
        "# Qui est-il ? Who is it?"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xXht4MOKHsLi"
      },
      "outputs": [],
      "source": [
        "# Download the file\n",
        "import pathlib\n",
        "\n",
        "path_to_zip = tf.keras.utils.get_file(\n",
        "    'fra-eng.zip', origin='http://storage.googleapis.com/download.tensorflow.org/data/fra-eng.zip',\n",
        "    extract=True)\n",
        "\n",
        "path_to_file = pathlib.Path(path_to_zip).parent/'fra.txt'"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "AHF6IPMoHt5l"
      },
      "outputs": [],
      "source": [
        "def load_data(path):\n",
        "    text = path.read_text(encoding='utf-8')\n",
        "    lines = text.splitlines()\n",
        "    pairs = [line.split('\\t') for line in lines]\n",
        "\n",
        "    inp = [inp for targ, inp in pairs]\n",
        "    targ = [targ for targ, inp in pairs]\n",
        "\n",
        "    return targ, inp"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UMlCRiBCHvW5",
        "outputId": "89c66d3b-eabd-4f61-9028-79ba5a65ef4f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "entrée :  Il est peut-être impossible d'obtenir un Corpus complètement dénué de fautes, étant donnée la nature de ce type d'entreprise collaborative. Cependant, si nous encourageons les membres à produire des phrases dans leurs propres langues plutôt que d'expérimenter dans les langues qu'ils apprennent, nous pourrions être en mesure de réduire les erreurs.\n"
          ]
        }
      ],
      "source": [
        "targ, inp = load_data(path_to_file)\n",
        "print(\"entrée : \",inp[-1])\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "NaGXKnwSHxNQ",
        "outputId": "99cef464-14f2-44ba-9d9a-81453198c4ed"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "cible :  It may be impossible to get a completely error-free corpus due to the nature of this kind of collaborative effort. However, if we encourage members to contribute sentences in their own languages rather than experiment in languages they are learning, we might be able to minimize errors.\n"
          ]
        }
      ],
      "source": [
        "print(\"cible : \",targ[-1])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "hBrHT_mMl6pQ",
        "outputId": "43be7a02-ba42-4147-eec3-ae0b80723aff"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "nombre de phrase contenu dans les données :  167130\n",
            "nombre de mot de la plus petite phrase :  4\n",
            "nombre de mot de la plus longue phrase :  349\n"
          ]
        }
      ],
      "source": [
        "# Taille des données en entrée\n",
        "print(\"nombre de phrase contenu dans les données : \", len(inp))\n",
        "# On entraine sur 167130 phrase avec des longueur variable allant des plus petites au plus longues\n",
        "\n",
        "# pour montrer qu'on a des longueurs différentes\n",
        "taille_input = [len(p) for p in inp]\n",
        "print(\"nombre de mot de la plus petite phrase : \", min(taille_input))\n",
        "print(\"nombre de mot de la plus longue phrase : \", max(taille_input))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "UNsS2Y61l6pR",
        "outputId": "8f00f820-b931-4350-8c3f-a2345114b768"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Si on prend toutes les données on aura 2611 exemples pour chaque epoque\n"
          ]
        }
      ],
      "source": [
        "BUFFER_SIZE = len(inp)\n",
        "BATCH_SIZE = 64\n",
        "print(\"Si on prend toutes les données on aura\",round(BUFFER_SIZE/BATCH_SIZE), \"exemples pour chaque epoque\")\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9eNLVLKLHyq4"
      },
      "outputs": [],
      "source": [
        "# on mélange les données pour ne pas avoir que des phrases de même longueur\n",
        "# pour un entrainement plus léger on peut limiter les données en entrée à (inp[:50000], targ[:50000]) \n",
        "dataset = tf.data.Dataset.from_tensor_slices((inp, targ)).shuffle(BUFFER_SIZE)\n",
        "dataset = dataset.batch(BATCH_SIZE)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "orxeToKgH0EJ",
        "outputId": "dfbd2bb8-49af-493c-eb89-5dbb952c59e8"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "tf.Tensor(\n",
            "[b'Qui \\xc3\\xaates-vous tous\\xe2\\x80\\xaf?'\n",
            " b\"Ce n'est pas ce que tu veux faire.\" b'Je connais votre genre.'\n",
            " b'Il vous a trahi.' b'Mon p\\xc3\\xa8re ne permettra pas cela.'], shape=(5,), dtype=string)\n",
            "\n",
            "tf.Tensor(\n",
            "[b'Who are you all?' b\"That's not what you want to do.\"\n",
            " b'I know your type.' b'He betrayed you.' b\"My father won't allow that.\"], shape=(5,), dtype=string)\n"
          ]
        }
      ],
      "source": [
        "for example_input_batch, example_target_batch in dataset.take(1):\n",
        "    print(example_input_batch[:5])\n",
        "    print()\n",
        "    print(example_target_batch[:5])\n",
        "    break"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hIGoSm71H1li",
        "outputId": "43e59ffb-f49c-4c2f-afd3-70dc45ed7061"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "b'Tu est toujours \\xc3\\xa0 la maison?'\n",
            "b'Tu est toujours a\\xcc\\x80 la maison?'\n"
          ]
        }
      ],
      "source": [
        "# On a des problème ou y'a des accents, il faut travailler en unicode, on normalise en montrant d'abord la lettre\n",
        "# La normalisation Unicode pour diviser les caractères accentués et remplacer les caractères de compatibilité par leurs équivalents ASCII.\n",
        "example_text = tf.constant('Tu est toujours à la maison?')\n",
        "\n",
        "print(example_text.numpy())\n",
        "print(tf_text.normalize_utf8(example_text, 'NFKD').numpy())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ddy3nG-rH4i3"
      },
      "outputs": [],
      "source": [
        "def tf_lower_and_split_punct(text):\n",
        "    # Split accecented characters.\n",
        "    text = tf_text.normalize_utf8(text, 'NFKD')\n",
        "    text = tf.strings.lower(text)\n",
        "    # Keep space, a to z, and select punctuation.\n",
        "    text = tf.strings.regex_replace(text, '[^ a-z.?!,¿]', '')\n",
        "    # Add spaces around punctuation.\n",
        "    text = tf.strings.regex_replace(text, '[.?!,¿]', r' \\0 ')\n",
        "    # Strip whitespace.\n",
        "    text = tf.strings.strip(text)\n",
        "    # on ajoute des indicateurs de début et fin de phrase\n",
        "    text = tf.strings.join(['[START]', text, '[END]'], separator=' ')\n",
        "    return text"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8YI0_5uWH5Sx",
        "outputId": "42ebb3b3-0397-45aa-9bdd-e6365ce673bb"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Tu est toujours à la maison?\n",
            "[START] tu est toujours a la maison ? [END]\n"
          ]
        }
      ],
      "source": [
        "print(example_text.numpy().decode())\n",
        "# on garde que la lettre sans accent\n",
        "print(tf_lower_and_split_punct(example_text).numpy().decode())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "zLhwT2JEH6j5"
      },
      "outputs": [],
      "source": [
        "max_vocab_size = 5000\n",
        "\n",
        "input_text_processor = tf.keras.layers.TextVectorization(\n",
        "    standardize=tf_lower_and_split_punct,\n",
        "    max_tokens=max_vocab_size)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "s7rJaX5YH8uw",
        "outputId": "ab253f2b-35df-48ba-8f73-5bb8ce51e303"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['', '[UNK]', '[START]', '[END]', '.', 'je', 'de', 'a', '?', 'pas']"
            ]
          },
          "execution_count": 19,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "input_text_processor.adapt(inp)\n",
        "\n",
        "# Here are the first 10 words from the vocabulary:\n",
        "input_text_processor.get_vocabulary()[:10]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6YvugutPH9k9",
        "outputId": "32aac95a-15ab-4dbb-8e48-fe63d8c1bfed"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['', '[UNK]', '[START]', '[END]', '.', 'i', 'you', 'to', 'the', '?']"
            ]
          },
          "execution_count": 20,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "output_text_processor = tf.keras.layers.TextVectorization(\n",
        "    standardize=tf_lower_and_split_punct,\n",
        "    max_tokens=max_vocab_size)\n",
        "\n",
        "output_text_processor.adapt(targ)\n",
        "output_text_processor.get_vocabulary()[:10]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "jz7y3diQH-6I",
        "outputId": "d87ebb0d-ba85-443f-f5a4-24c36ec83d6e"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(3, 10), dtype=int64, numpy=\n",
              "array([[  2,  39, 105,  85,   8,   3,   0,   0,   0,   0],\n",
              "       [  2,  20,  70,   9,  20,  10,  21,  47,  31,   4],\n",
              "       [  2,   5, 372,  59, 523,   4,   3,   0,   0,   0]])>"
            ]
          },
          "execution_count": 21,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "example_tokens = input_text_processor(example_input_batch)\n",
        "# quelques phrases tokenisées\n",
        "example_tokens[:3, :10]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "Kt20FiD0IBJM",
        "outputId": "6a7693e3-552f-4b32-c428-3f401ebe5a7e"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "'[START] qui etesvous tous ? [END]             '"
            ]
          },
          "execution_count": 22,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "input_vocab = np.array(input_text_processor.get_vocabulary())\n",
        "tokens = input_vocab[example_tokens[0].numpy()]\n",
        "' '.join(tokens)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "76MHMR-fIBPk",
        "outputId": "5dca77b9-64d8-4bff-f1b6-371c9e6f2bc6"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Mask')"
            ]
          },
          "execution_count": 23,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.subplot(1, 2, 1)\n",
        "plt.pcolormesh(example_tokens)\n",
        "plt.title('Token IDs')\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.pcolormesh(example_tokens != 0)\n",
        "plt.title('Mask')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "svEnv8qWIBXW"
      },
      "outputs": [],
      "source": [
        "embedding_dim = 256\n",
        "units = 1000 # nombre de neurones des couches de l'encodeur et décodeur"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "k91qJdrYIB1l"
      },
      "outputs": [],
      "source": [
        "# Till now, we have used a monodirectional GRU instead of a bidirectional as first tests.\n",
        "# However, in the next version we will use the bidirectional GRU.\n",
        "class Encoder(tf.keras.layers.Layer):\n",
        "    def __init__(self, input_vocab_size, embedding_dim, enc_units):\n",
        "        super(Encoder, self).__init__()\n",
        "        self.enc_units = enc_units\n",
        "        self.input_vocab_size = input_vocab_size\n",
        "\n",
        "        # The embedding layer converts tokens to vectors\n",
        "        self.embedding = tf.keras.layers.Embedding(self.input_vocab_size,\n",
        "                                                   embedding_dim)\n",
        "\n",
        "        # The GRU RNN layer processes those vectors sequentially.\n",
        "        self.gru = tf.keras.layers.GRU(self.enc_units,\n",
        "                                       # Return the sequence and state\n",
        "                                       return_sequences=True,\n",
        "                                       return_state=True,\n",
        "                                       recurrent_initializer='glorot_uniform')\n",
        "\n",
        "    def call(self, tokens, state=None):\n",
        "        shape_checker = ShapeChecker()\n",
        "        shape_checker(tokens, ('batch', 's'))\n",
        "\n",
        "        # 2. The embedding layer looks up the embedding for each token.\n",
        "        vectors = self.embedding(tokens)\n",
        "        shape_checker(vectors, ('batch', 's', 'embed_dim'))\n",
        "\n",
        "        # 3. The GRU processes the embedding sequence.\n",
        "        #    output shape: (batch, s, enc_units)\n",
        "        #    state shape: (batch, enc_units)\n",
        "        output, state = self.gru(vectors, initial_state=state)\n",
        "        shape_checker(output, ('batch', 's', 'enc_units'))\n",
        "        shape_checker(state, ('batch', 'enc_units'))\n",
        "\n",
        "        # 4. Returns the new sequence and its state.\n",
        "        return output, state"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2CWuL_sTIH7f",
        "outputId": "c1112240-8c40-4da7-fd19-a8c7997d0664"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Input batch, shape (batch): (64,)\n",
            "Input batch tokens, shape (batch, s): (64, 19)\n",
            "Encoder output, shape (batch, s, units): (64, 19, 1000)\n",
            "Encoder state, shape (batch, units): (64, 1000)\n"
          ]
        }
      ],
      "source": [
        "# Convert the input text to tokens.\n",
        "example_tokens = input_text_processor(example_input_batch)\n",
        "\n",
        "# Encode the input sequence.\n",
        "encoder = Encoder(input_text_processor.vocabulary_size(),\n",
        "                  embedding_dim, units)\n",
        "example_enc_output, example_enc_state = encoder(example_tokens)\n",
        "\n",
        "print(f'Input batch, shape (batch): {example_input_batch.shape}')\n",
        "print(f'Input batch tokens, shape (batch, s): {example_tokens.shape}')\n",
        "print(f'Encoder output, shape (batch, s, units): {example_enc_output.shape}')\n",
        "print(f'Encoder state, shape (batch, units): {example_enc_state.shape}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "sgRPoDf0IIAr"
      },
      "outputs": [],
      "source": [
        "class BahdanauAttention(tf.keras.layers.Layer):\n",
        "    def __init__(self, units):\n",
        "        super().__init__()\n",
        "        # For Eqn. (4), the  Bahdanau attention\n",
        "        self.W1 = tf.keras.layers.Dense(units, use_bias=False)\n",
        "        self.W2 = tf.keras.layers.Dense(units, use_bias=False)\n",
        "\n",
        "        self.attention = tf.keras.layers.AdditiveAttention()\n",
        "\n",
        "    def call(self, query, value, mask):\n",
        "        shape_checker = ShapeChecker()\n",
        "        shape_checker(query, ('batch', 't', 'query_units'))\n",
        "        shape_checker(value, ('batch', 's', 'value_units'))\n",
        "        shape_checker(mask, ('batch', 's'))\n",
        "\n",
        "        # From Eqn. (4), `W1@ht`.\n",
        "        w1_query = self.W1(query)\n",
        "        shape_checker(w1_query, ('batch', 't', 'attn_units'))\n",
        "\n",
        "        # From Eqn. (4), `W2@hs`.\n",
        "        w2_key = self.W2(value)\n",
        "        shape_checker(w2_key, ('batch', 's', 'attn_units'))\n",
        "\n",
        "        query_mask = tf.ones(tf.shape(query)[:-1], dtype=bool)\n",
        "        value_mask = mask\n",
        "\n",
        "        context_vector, attention_weights = self.attention(\n",
        "            inputs = [w1_query, value, w2_key],\n",
        "            mask=[query_mask, value_mask],\n",
        "            return_attention_scores = True,\n",
        "        )\n",
        "        shape_checker(context_vector, ('batch', 't', 'value_units'))\n",
        "        shape_checker(attention_weights, ('batch', 't', 's'))\n",
        "\n",
        "        return context_vector, attention_weights"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "kvktv7CgIIDJ"
      },
      "outputs": [],
      "source": [
        "attention_layer = BahdanauAttention(units)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0GaMTu6-IIGy",
        "outputId": "9c34983c-7d6c-4902-8d7e-36669d946e89"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "TensorShape([64, 19])"
            ]
          },
          "execution_count": 29,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "(example_tokens != 0).shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PWRqXkQ1IIPp",
        "outputId": "a452eb05-6f13-4ece-fc36-2d549b948af5"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Attention result shape: (batch_size, query_seq_length, units):           (64, 2, 1000)\n",
            "Attention weights shape: (batch_size, query_seq_length, value_seq_length): (64, 2, 19)\n"
          ]
        }
      ],
      "source": [
        "# Later, the decoder will generate this attention query\n",
        "example_attention_query = tf.random.normal(shape=[len(example_tokens), 2, 10])\n",
        "\n",
        "# Attend to the encoded tokens\n",
        "\n",
        "context_vector, attention_weights = attention_layer(\n",
        "    query=example_attention_query,\n",
        "    value=example_enc_output,\n",
        "    mask=(example_tokens != 0))\n",
        "\n",
        "print(f'Attention result shape: (batch_size, query_seq_length, units):           {context_vector.shape}')\n",
        "print(f'Attention weights shape: (batch_size, query_seq_length, value_seq_length): {attention_weights.shape}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 298
        },
        "id": "UVwcPJYEIITF",
        "outputId": "fd325d04-47f5-426b-cb2b-5b2b0aced9b9"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Mask')"
            ]
          },
          "execution_count": 31,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 2 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.subplot(1, 2, 1)\n",
        "plt.pcolormesh(attention_weights[:, 0, :])\n",
        "plt.title('Attention weights')\n",
        "\n",
        "plt.subplot(1, 2, 2)\n",
        "plt.pcolormesh(example_tokens != 0)\n",
        "plt.title('Mask')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Mkm6ET0wISBG",
        "outputId": "61f0f428-2fb8-414b-ef21-4a9b402760a6"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "TensorShape([64, 2, 19])"
            ]
          },
          "execution_count": 32,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "attention_weights.shape"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6HshcWeJISDy"
      },
      "outputs": [],
      "source": [
        "attention_slice = attention_weights[0, 0].numpy()\n",
        "attention_slice = attention_slice[attention_slice != 0]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "gbWSszPKISGY"
      },
      "outputs": [],
      "source": [
        "class Decoder(tf.keras.layers.Layer):\n",
        "    def __init__(self, output_vocab_size, embedding_dim, dec_units):\n",
        "        super(Decoder, self).__init__()\n",
        "        self.dec_units = dec_units\n",
        "        self.output_vocab_size = output_vocab_size\n",
        "        self.embedding_dim = embedding_dim\n",
        "\n",
        "        # For Step 1. The embedding layer convets token IDs to vectors\n",
        "        self.embedding = tf.keras.layers.Embedding(self.output_vocab_size,\n",
        "                                                   embedding_dim)\n",
        "\n",
        "        # For Step 2. The RNN keeps track of what's been generated so far.\n",
        "        self.gru = tf.keras.layers.GRU(self.dec_units,\n",
        "                                       return_sequences=True,\n",
        "                                       return_state=True,\n",
        "                                       recurrent_initializer='glorot_uniform')\n",
        "\n",
        "        # For step 3. The RNN output will be the query for the attention layer.\n",
        "        self.attention = BahdanauAttention(self.dec_units)\n",
        "\n",
        "        # For step 4. Eqn. (3): converting `ct` to `at`\n",
        "        self.Wc = tf.keras.layers.Dense(dec_units, activation=tf.math.tanh,\n",
        "                                        use_bias=False)\n",
        "\n",
        "        # For step 5. This fully connected layer produces the logits for each\n",
        "        # output token.\n",
        "        self.fc = tf.keras.layers.Dense(self.output_vocab_size)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ru5MZzPuISJG"
      },
      "outputs": [],
      "source": [
        "class DecoderInput(typing.NamedTuple):\n",
        "    new_tokens: Any\n",
        "    enc_output: Any\n",
        "    mask: Any\n",
        "\n",
        "class DecoderOutput(typing.NamedTuple):\n",
        "    logits: Any\n",
        "    attention_weights: Any"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9Pr6bAwsISMV"
      },
      "outputs": [],
      "source": [
        "def call(self,\n",
        "         inputs: DecoderInput,\n",
        "         state=None) -> Tuple[DecoderOutput, tf.Tensor]:\n",
        "    shape_checker = ShapeChecker()\n",
        "    shape_checker(inputs.new_tokens, ('batch', 't'))\n",
        "    shape_checker(inputs.enc_output, ('batch', 's', 'enc_units'))\n",
        "    shape_checker(inputs.mask, ('batch', 's'))\n",
        "\n",
        "    if state is not None:\n",
        "        shape_checker(state, ('batch', 'dec_units'))\n",
        "\n",
        "    # Step 1. Lookup the embeddings\n",
        "    vectors = self.embedding(inputs.new_tokens)\n",
        "    shape_checker(vectors, ('batch', 't', 'embedding_dim'))\n",
        "\n",
        "    # Step 2. Process one step with the RNN\n",
        "    rnn_output, state = self.gru(vectors, initial_state=state)\n",
        "\n",
        "    shape_checker(rnn_output, ('batch', 't', 'dec_units'))\n",
        "    shape_checker(state, ('batch', 'dec_units'))\n",
        "\n",
        "    # Step 3. Use the RNN output as the query for the attention over the\n",
        "    # encoder output.\n",
        "    context_vector, attention_weights = self.attention(\n",
        "        query=rnn_output, value=inputs.enc_output, mask=inputs.mask)\n",
        "    shape_checker(context_vector, ('batch', 't', 'dec_units'))\n",
        "    shape_checker(attention_weights, ('batch', 't', 's'))\n",
        "\n",
        "    # Step 4. Eqn. (3): Join the context_vector and rnn_output\n",
        "    #     [ct; ht] shape: (batch t, value_units + query_units)\n",
        "    context_and_rnn_output = tf.concat([context_vector, rnn_output], axis=-1)\n",
        "\n",
        "    # Step 4. Eqn. (3): `at = tanh(Wc@[ct; ht])`\n",
        "    attention_vector = self.Wc(context_and_rnn_output)\n",
        "    shape_checker(attention_vector, ('batch', 't', 'dec_units'))\n",
        "\n",
        "    # Step 5. Generate logit predictions:\n",
        "    logits = self.fc(attention_vector)\n",
        "    shape_checker(logits, ('batch', 't', 'output_vocab_size'))\n",
        "\n",
        "    return DecoderOutput(logits, attention_weights), state"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "iK6mVNliIYwg"
      },
      "outputs": [],
      "source": [
        "Decoder.call = call"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Jb3DqDSpIY36"
      },
      "outputs": [],
      "source": [
        "decoder = Decoder(output_text_processor.vocabulary_size(),\n",
        "                  embedding_dim, units)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "_j9WT0nLIY9Q"
      },
      "outputs": [],
      "source": [
        "# Convert the target sequence, and collect the \"[START]\" tokens\n",
        "example_output_tokens = output_text_processor(example_target_batch)\n",
        "\n",
        "start_index = output_text_processor.get_vocabulary().index('[START]')\n",
        "first_token = tf.constant([[start_index]] * example_output_tokens.shape[0])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8TS8n2mNIZAE",
        "outputId": "6b4164c8-62b7-49e4-8549-21e8d612a03c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "logits shape: (batch_size, t, output_vocab_size) (64, 1, 5000)\n",
            "state shape: (batch_size, dec_units) (64, 1000)\n"
          ]
        }
      ],
      "source": [
        "# Run the decoder\n",
        "dec_result, dec_state = decoder(\n",
        "    inputs = DecoderInput(new_tokens=first_token,\n",
        "                          enc_output=example_enc_output,\n",
        "                          mask=(example_tokens != 0)),\n",
        "    state = example_enc_state\n",
        ")\n",
        "\n",
        "print(f'logits shape: (batch_size, t, output_vocab_size) {dec_result.logits.shape}')\n",
        "print(f'state shape: (batch_size, dec_units) {dec_state.shape}')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "M_PcMNIbIdnb"
      },
      "outputs": [],
      "source": [
        "sampled_token = tf.random.categorical(dec_result.logits[:, 0, :], num_samples=1)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lvk-N4ADIdp_",
        "outputId": "562f1988-0452-4308-86ab-be6718792ec0"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([['suspected'],\n",
              "       ['plan'],\n",
              "       ['dozed'],\n",
              "       ['way'],\n",
              "       ['caused']], dtype='<U16')"
            ]
          },
          "execution_count": 42,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "vocab = np.array(output_text_processor.get_vocabulary())\n",
        "first_word = vocab[sampled_token.numpy()]\n",
        "first_word[:5]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1Y9W7LfcIdsl"
      },
      "outputs": [],
      "source": [
        "dec_result, dec_state = decoder(\n",
        "    DecoderInput(sampled_token,\n",
        "                 example_enc_output,\n",
        "                 mask=(example_tokens != 0)),\n",
        "    state=dec_state)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JCYD8NLIIdvY"
      },
      "outputs": [],
      "source": [
        "class MaskedLoss(tf.keras.losses.Loss):\n",
        "    def __init__(self):\n",
        "        self.name = 'masked_loss'\n",
        "        self.loss = tf.keras.losses.SparseCategoricalCrossentropy(\n",
        "            from_logits=True, reduction='none')\n",
        "\n",
        "    def __call__(self, y_true, y_pred):\n",
        "        shape_checker = ShapeChecker()\n",
        "        shape_checker(y_true, ('batch', 't'))\n",
        "        shape_checker(y_pred, ('batch', 't', 'logits'))\n",
        "\n",
        "        # Calculate the loss for each item in the batch.\n",
        "        loss = self.loss(y_true, y_pred)\n",
        "        shape_checker(loss, ('batch', 't'))\n",
        "\n",
        "        # Mask off the losses on padding.\n",
        "        mask = tf.cast(y_true != 0, tf.float32)\n",
        "        shape_checker(mask, ('batch', 't'))\n",
        "        loss *= mask\n",
        "\n",
        "        # Return the total.\n",
        "        return tf.reduce_sum(loss)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "vOpEjv6WIdxr"
      },
      "outputs": [],
      "source": [
        "class TrainTranslator(tf.keras.Model):\n",
        "    def __init__(self, embedding_dim, units,\n",
        "                   input_text_processor,\n",
        "                   output_text_processor, \n",
        "                   use_tf_function=True):\n",
        "        super().__init__()\n",
        "        # Build the encoder and decoder\n",
        "        encoder = Encoder(input_text_processor.vocabulary_size(),\n",
        "                          embedding_dim, units)\n",
        "        decoder = Decoder(output_text_processor.vocabulary_size(),\n",
        "                          embedding_dim, units)\n",
        "\n",
        "        self.encoder = encoder\n",
        "        self.decoder = decoder\n",
        "        self.input_text_processor = input_text_processor\n",
        "        self.output_text_processor = output_text_processor\n",
        "        self.use_tf_function = use_tf_function\n",
        "        self.shape_checker = ShapeChecker()\n",
        "\n",
        "    def train_step(self, inputs):\n",
        "        self.shape_checker = ShapeChecker()\n",
        "        if self.use_tf_function:\n",
        "            return self._tf_train_step(inputs)\n",
        "        else:\n",
        "            return self._train_step(inputs)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "sx8DzYd4Id1F"
      },
      "outputs": [],
      "source": [
        "def _preprocess(self, input_text, target_text):\n",
        "    self.shape_checker(input_text, ('batch',))\n",
        "    self.shape_checker(target_text, ('batch',))\n",
        "\n",
        "    # Convert the text to token IDs\n",
        "    input_tokens = self.input_text_processor(input_text)\n",
        "    target_tokens = self.output_text_processor(target_text)\n",
        "    self.shape_checker(input_tokens, ('batch', 's'))\n",
        "    self.shape_checker(target_tokens, ('batch', 't'))\n",
        "\n",
        "    # Convert IDs to masks.\n",
        "    input_mask = input_tokens != 0\n",
        "    self.shape_checker(input_mask, ('batch', 's'))\n",
        "\n",
        "    target_mask = target_tokens != 0\n",
        "    self.shape_checker(target_mask, ('batch', 't'))\n",
        "\n",
        "    return input_tokens, input_mask, target_tokens, target_mask"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "12vHRtMxIZDe"
      },
      "outputs": [],
      "source": [
        "TrainTranslator._preprocess = _preprocess"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1YbqjLtFImSG"
      },
      "outputs": [],
      "source": [
        "def _train_step(self, inputs):\n",
        "    input_text, target_text = inputs  \n",
        "\n",
        "    (input_tokens, input_mask,\n",
        "    target_tokens, target_mask) = self._preprocess(input_text, target_text)\n",
        "\n",
        "    max_target_length = tf.shape(target_tokens)[1]\n",
        "\n",
        "    with tf.GradientTape() as tape:\n",
        "        # Encode the input\n",
        "        enc_output, enc_state = self.encoder(input_tokens)\n",
        "        self.shape_checker(enc_output, ('batch', 's', 'enc_units'))\n",
        "        self.shape_checker(enc_state, ('batch', 'enc_units'))\n",
        "\n",
        "        # Initialize the decoder's state to the encoder's final state.\n",
        "        # This only works if the encoder and decoder have the same number of\n",
        "        # units.\n",
        "        dec_state = enc_state\n",
        "        loss = tf.constant(0.0)\n",
        "\n",
        "        for t in tf.range(max_target_length-1):\n",
        "            # Pass in two tokens from the target sequence:\n",
        "            # 1. The current input to the decoder.\n",
        "            # 2. The target for the decoder's next prediction.\n",
        "            new_tokens = target_tokens[:, t:t+2]\n",
        "            step_loss, dec_state = self._loop_step(new_tokens, input_mask,\n",
        "                                                 enc_output, dec_state)\n",
        "            loss = loss + step_loss\n",
        "\n",
        "        # Average the loss over all non padding tokens.\n",
        "        average_loss = loss / tf.reduce_sum(tf.cast(target_mask, tf.float32))\n",
        "\n",
        "    # Apply an optimization step\n",
        "    variables = self.trainable_variables \n",
        "    gradients = tape.gradient(average_loss, variables)\n",
        "    self.optimizer.apply_gradients(zip(gradients, variables))\n",
        "\n",
        "    # Return a dict mapping metric names to current value\n",
        "    return {'batch_loss': average_loss}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1KDzgG0UImUz"
      },
      "outputs": [],
      "source": [
        "TrainTranslator._train_step = _train_step"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "N8OoM2KAImYS"
      },
      "outputs": [],
      "source": [
        "def _loop_step(self, new_tokens, input_mask, enc_output, dec_state):\n",
        "    input_token, target_token = new_tokens[:, 0:1], new_tokens[:, 1:2]\n",
        "\n",
        "    # Run the decoder one step.\n",
        "    decoder_input = DecoderInput(new_tokens=input_token,\n",
        "                               enc_output=enc_output,\n",
        "                               mask=input_mask)\n",
        "\n",
        "    dec_result, dec_state = self.decoder(decoder_input, state=dec_state)\n",
        "    self.shape_checker(dec_result.logits, ('batch', 't1', 'logits'))\n",
        "    self.shape_checker(dec_result.attention_weights, ('batch', 't1', 's'))\n",
        "    self.shape_checker(dec_state, ('batch', 'dec_units'))\n",
        "\n",
        "    # `self.loss` returns the total for non-padded tokens\n",
        "    y = target_token\n",
        "    y_pred = dec_result.logits\n",
        "    step_loss = self.loss(y, y_pred)\n",
        "\n",
        "    return step_loss, dec_state"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "9W-l9eI0Imlk"
      },
      "outputs": [],
      "source": [
        "TrainTranslator._loop_step = _loop_step"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "lCdnYuT1ItjR"
      },
      "outputs": [],
      "source": [
        "translator = TrainTranslator(\n",
        "    embedding_dim, units,\n",
        "    input_text_processor=input_text_processor,\n",
        "    output_text_processor=output_text_processor,\n",
        "    use_tf_function=False)\n",
        "\n",
        "# Configure the loss and optimizer\n",
        "translator.compile(\n",
        "    optimizer=tf.optimizers.Adam(),\n",
        "    loss=MaskedLoss(),\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8fvyimEZItl2",
        "outputId": "4d82f9ae-7ce7-46ec-8266-769dd8d3ed74"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "8.517193191416238"
            ]
          },
          "execution_count": 53,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "np.log(output_text_processor.vocabulary_size())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bI-SiuJmItof",
        "outputId": "28ff017c-3339-4f97-ebf4-95b9747877b6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=7.5706186>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=7.53916>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=7.4789295>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=7.3034635>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=6.647581>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=5.1521173>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=4.648046>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=4.4589624>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=4.2081137>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=4.0415454>}\n",
            "\n",
            "CPU times: user 1min 56s, sys: 2.41 s, total: 1min 58s\n",
            "Wall time: 44.5 s\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "for n in range(10):\n",
        "      print(translator.train_step([example_input_batch, example_target_batch]))\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Z21BnHVaItq7"
      },
      "outputs": [],
      "source": [
        "@tf.function(input_signature=[[tf.TensorSpec(dtype=tf.string, shape=[None]),\n",
        "                               tf.TensorSpec(dtype=tf.string, shape=[None])]])\n",
        "def _tf_train_step(self, inputs):\n",
        "      return self._train_step(inputs)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "MXdGwpIQIyNY"
      },
      "outputs": [],
      "source": [
        "TrainTranslator._tf_train_step = _tf_train_step"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "BWfspKE6IyXG"
      },
      "outputs": [],
      "source": [
        "translator.use_tf_function = True"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "6PdCoV1eIyZ2",
        "outputId": "5468bdaf-cfa1-4394-e853-808dea14e362"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.9242303>}"
            ]
          },
          "execution_count": 58,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "translator.train_step([example_input_batch, example_target_batch])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "eCIE07YkIycs",
        "outputId": "149ae27b-9990-46b6-a922-e92367cd2f32"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.8988004>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.9063084>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.8385468>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.7314546>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.7027087>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.6707034>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.617001>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.5801218>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.5387888>}\n",
            "{'batch_loss': <tf.Tensor: shape=(), dtype=float32, numpy=3.4983494>}\n",
            "\n",
            "CPU times: user 1min 7s, sys: 2.23 s, total: 1min 9s\n",
            "Wall time: 17.8 s\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "for n in range(10):\n",
        "    print(translator.train_step([example_input_batch, example_target_batch]))\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 320
        },
        "id": "HTNg6SLlIyex",
        "outputId": "8b391631-e61b-468b-b078-6570de4eec41"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "....................................................................................................\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "[<matplotlib.lines.Line2D at 0x7fe01dea4d60>]"
            ]
          },
          "execution_count": 60,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "losses = []\n",
        "for n in range(100):\n",
        "    print('.', end='')\n",
        "    logs = translator.train_step([example_input_batch, example_target_batch])\n",
        "    losses.append(logs['batch_loss'].numpy())\n",
        "\n",
        "print()\n",
        "plt.plot(losses)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ZOob3Y7iIyh2"
      },
      "outputs": [],
      "source": [
        "train_translator = TrainTranslator(\n",
        "    embedding_dim, units,\n",
        "    input_text_processor=input_text_processor,\n",
        "    output_text_processor=output_text_processor)\n",
        "\n",
        "# Configure the loss and optimizer\n",
        "train_translator.compile(\n",
        "    optimizer=tf.optimizers.Adam(),\n",
        "    loss=MaskedLoss(),\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "6W-hnB_iI5LK"
      },
      "outputs": [],
      "source": [
        "class BatchLogs(tf.keras.callbacks.Callback):\n",
        "    def __init__(self, key):\n",
        "        self.key = key\n",
        "        self.logs = []\n",
        "\n",
        "    def on_train_batch_end(self, n, logs):\n",
        "        self.logs.append(logs[self.key])\n",
        "\n",
        "batch_loss = BatchLogs('batch_loss')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "UcHjoTAlI5Un",
        "outputId": "2c4e08c2-11bd-4c8b-e4ce-9a08c7dafb8f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/3\n",
            "2612/2612 [==============================] - 4272s 2s/step - batch_loss: 1.7486\n",
            "Epoch 2/3\n",
            "2612/2612 [==============================] - 4253s 2s/step - batch_loss: 0.9014\n",
            "Epoch 3/3\n",
            "2612/2612 [==============================] - 4269s 2s/step - batch_loss: 0.6916\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7f125f676d40>"
            ]
          },
          "execution_count": 63,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "train_translator.fit(dataset, epochs=3,\n",
        "                     callbacks=[batch_loss])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "KWSGNjXGI5Xg",
        "outputId": "03a4f37a-fc80-44b3-dacf-485d60e2833f"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Text(0, 0.5, 'CE/token')"
            ]
          },
          "execution_count": 64,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.plot(batch_loss.logs)\n",
        "plt.ylim([0, 3])\n",
        "plt.xlabel('Batch #')\n",
        "plt.ylabel('CE/token')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yjs7AKaGI5bD"
      },
      "outputs": [],
      "source": [
        "class Translator(tf.Module):\n",
        "\n",
        "    def __init__(self, encoder, decoder, input_text_processor,\n",
        "                   output_text_processor):\n",
        "        self.encoder = encoder\n",
        "        self.decoder = decoder\n",
        "        self.input_text_processor = input_text_processor\n",
        "        self.output_text_processor = output_text_processor\n",
        "\n",
        "        self.output_token_string_from_index = (\n",
        "            tf.keras.layers.StringLookup(\n",
        "                vocabulary=output_text_processor.get_vocabulary(),\n",
        "                mask_token='',\n",
        "                invert=True))\n",
        "\n",
        "        # The output should never generate padding, unknown, or start.\n",
        "        index_from_string = tf.keras.layers.StringLookup(\n",
        "            vocabulary=output_text_processor.get_vocabulary(), mask_token='')\n",
        "        token_mask_ids = index_from_string(['', '[UNK]', '[START]']).numpy()\n",
        "\n",
        "        token_mask = np.zeros([index_from_string.vocabulary_size()], dtype=np.bool)\n",
        "        token_mask[np.array(token_mask_ids)] = True\n",
        "        self.token_mask = token_mask\n",
        "\n",
        "        self.start_token = index_from_string(tf.constant('[START]'))\n",
        "        self.end_token = index_from_string(tf.constant('[END]'))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "rUA3hSxCItuO",
        "outputId": "99989381-91ca-4fc1-df7e-21d347c0cbe4"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/tmp/wsuser/ipykernel_416/2407734992.py:21: DeprecationWarning: `np.bool` is a deprecated alias for the builtin `bool`. To silence this warning, use `bool` by itself. Doing this will not modify any behavior and is safe. If you specifically wanted the numpy scalar type, use `np.bool_` here.\n",
            "Deprecated in NumPy 1.20; for more details and guidance: https://numpy.org/devdocs/release/1.20.0-notes.html#deprecations\n",
            "  token_mask = np.zeros([index_from_string.vocabulary_size()], dtype=np.bool)\n"
          ]
        }
      ],
      "source": [
        "translator = Translator(\n",
        "    encoder=train_translator.encoder,\n",
        "    decoder=train_translator.decoder,\n",
        "    input_text_processor=input_text_processor,\n",
        "    output_text_processor=output_text_processor,\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "0CVvsih1I_tM"
      },
      "outputs": [],
      "source": [
        "def tokens_to_text(self, result_tokens):\n",
        "    shape_checker = ShapeChecker()\n",
        "    shape_checker(result_tokens, ('batch', 't'))\n",
        "    result_text_tokens = self.output_token_string_from_index(result_tokens)\n",
        "    shape_checker(result_text_tokens, ('batch', 't'))\n",
        "\n",
        "    result_text = tf.strings.reduce_join(result_text_tokens,\n",
        "                                       axis=1, separator=' ')\n",
        "    shape_checker(result_text, ('batch'))\n",
        "\n",
        "    result_text = tf.strings.strip(result_text)\n",
        "    shape_checker(result_text, ('batch',))\n",
        "    return result_text"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1nh-e-SXI_vg"
      },
      "outputs": [],
      "source": [
        "Translator.tokens_to_text = tokens_to_text"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "zO0Z4KbaI_yC",
        "outputId": "6602a710-e209-4821-c0c8-a43209c6607b"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([b'choose terrible', b'rate resembles', b'due holidays',\n",
              "       b'row alphabet', b'dropped grades'], dtype=object)"
            ]
          },
          "execution_count": 69,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "example_output_tokens = tf.random.uniform(\n",
        "    shape=[5, 2], minval=0, dtype=tf.int64,\n",
        "    maxval=output_text_processor.vocabulary_size())\n",
        "translator.tokens_to_text(example_output_tokens).numpy()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "DJz3zSEuI_0m"
      },
      "outputs": [],
      "source": [
        "def sample(self, logits, temperature):\n",
        "    shape_checker = ShapeChecker()\n",
        "    # 't' is usually 1 here.\n",
        "    shape_checker(logits, ('batch', 't', 'vocab'))\n",
        "    shape_checker(self.token_mask, ('vocab',))\n",
        "\n",
        "    token_mask = self.token_mask[tf.newaxis, tf.newaxis, :]\n",
        "    shape_checker(token_mask, ('batch', 't', 'vocab'), broadcast=True)\n",
        "\n",
        "    # Set the logits for all masked tokens to -inf, so they are never chosen.\n",
        "    logits = tf.where(self.token_mask, -np.inf, logits)\n",
        "\n",
        "    if temperature == 0.0:\n",
        "        new_tokens = tf.argmax(logits, axis=-1)\n",
        "    else: \n",
        "        logits = tf.squeeze(logits, axis=1)\n",
        "        new_tokens = tf.random.categorical(logits/temperature,\n",
        "                                        num_samples=1)\n",
        "\n",
        "    shape_checker(new_tokens, ('batch', 't'))\n",
        "\n",
        "    return new_tokens"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "2gBCRkaSI_4J"
      },
      "outputs": [],
      "source": [
        "Translator.sample = sample"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "TVTu3AF0JEc0",
        "outputId": "119bf26b-6714-4b9f-8e0e-0b62ef849e46"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(5, 1), dtype=int64, numpy=\n",
              "array([[2228],\n",
              "       [3527],\n",
              "       [2962],\n",
              "       [2177],\n",
              "       [4341]])>"
            ]
          },
          "execution_count": 72,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "example_logits = tf.random.normal([5, 1, output_text_processor.vocabulary_size()])\n",
        "example_output_tokens = translator.sample(example_logits, temperature=1.0)\n",
        "example_output_tokens"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "YVfvmDi_JEfT"
      },
      "outputs": [],
      "source": [
        "def translate_unrolled(self,\n",
        "                       input_text, *,\n",
        "                       max_length=50,\n",
        "                       return_attention=True,\n",
        "                       temperature=1.0):\n",
        "    batch_size = tf.shape(input_text)[0]\n",
        "    input_tokens = self.input_text_processor(input_text)\n",
        "    enc_output, enc_state = self.encoder(input_tokens)\n",
        "\n",
        "    dec_state = enc_state\n",
        "    new_tokens = tf.fill([batch_size, 1], self.start_token)\n",
        "\n",
        "    result_tokens = []\n",
        "    attention = []\n",
        "    done = tf.zeros([batch_size, 1], dtype=tf.bool)\n",
        "\n",
        "    for _ in range(max_length):\n",
        "        dec_input = DecoderInput(new_tokens=new_tokens,\n",
        "                                 enc_output=enc_output,\n",
        "                                 mask=(input_tokens!=0))\n",
        "\n",
        "        dec_result, dec_state = self.decoder(dec_input, state=dec_state)\n",
        "\n",
        "        attention.append(dec_result.attention_weights)\n",
        "\n",
        "        new_tokens = self.sample(dec_result.logits, temperature)\n",
        "\n",
        "        # If a sequence produces an `end_token`, set it `done`\n",
        "        done = done | (new_tokens == self.end_token)\n",
        "        # Once a sequence is done it only produces 0-padding.\n",
        "        new_tokens = tf.where(done, tf.constant(0, dtype=tf.int64), new_tokens)\n",
        "\n",
        "        # Collect the generated tokens\n",
        "        result_tokens.append(new_tokens)\n",
        "\n",
        "        if tf.executing_eagerly() and tf.reduce_all(done):\n",
        "            break\n",
        "\n",
        "    # Convert the list of generates token ids to a list of strings.\n",
        "    result_tokens = tf.concat(result_tokens, axis=-1)\n",
        "    result_text = self.tokens_to_text(result_tokens)\n",
        "\n",
        "    if return_attention:\n",
        "        attention_stack = tf.concat(attention, axis=1)\n",
        "        return {'text': result_text, 'attention': attention_stack}\n",
        "    else:\n",
        "        return {'text': result_text}"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Bl_yoRDtJEhi"
      },
      "outputs": [],
      "source": [
        "Translator.translate = translate_unrolled"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FQ1yNTIul6pr",
        "outputId": "f7c1f7b5-65b5-4e27-f89f-bbd1ac85193c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Nous sommes fortes.\n"
          ]
        }
      ],
      "source": [
        "print(inp[4500])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JD_PzuD1JEkL",
        "outputId": "2af7fce0-4877-4707-9678-4e83d4e525f4"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "itll be place tomorrow .\n",
            "were strong .\n",
            "\n",
            "CPU times: user 274 ms, sys: 12.1 ms, total: 286 ms\n",
            "Wall time: 188 ms\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "input_text = tf.constant([\n",
        "    'Ça aura lieu demain.', # \"It's really cold here.\"\n",
        "    'Nous sommes fortes.', # \"This is my life.\"\"\n",
        "])\n",
        "\n",
        "result = translator.translate(\n",
        "    input_text = input_text)\n",
        "\n",
        "print(result['text'][0].numpy().decode())\n",
        "print(result['text'][1].numpy().decode())\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "mPSv4nLeJLWW"
      },
      "outputs": [],
      "source": [
        "@tf.function(input_signature=[tf.TensorSpec(dtype=tf.string, shape=[None])])\n",
        "def tf_translate(self, input_text):\n",
        "      return self.translate(input_text)\n",
        "\n",
        "Translator.tf_translate = tf_translate"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "c4JCqFikJLZo",
        "outputId": "b97de634-b70e-40bf-b214-4bcf59bd3b8d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "CPU times: user 18.7 s, sys: 113 ms, total: 18.9 s\n",
            "Wall time: 18.2 s\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "result = translator.tf_translate(\n",
        "    input_text = input_text)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Y_ARMNWSJLbu",
        "outputId": "d28ea04d-ab87-4fbb-9c00-0c5588de883b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "its going to be place tomorrow .\n",
            "were strong .\n",
            "\n",
            "CPU times: user 881 ms, sys: 45.4 ms, total: 926 ms\n",
            "Wall time: 317 ms\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "result = translator.tf_translate(\n",
        "    input_text = input_text)\n",
        "\n",
        "print(result['text'][0].numpy().decode())\n",
        "print(result['text'][1].numpy().decode())\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "tJhJaIYAJLhD",
        "outputId": "aaa92ccc-ebb6-4c4a-a0cf-bc7c1c37a134"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "itll be here tomorrow .\n",
            "were strong .\n",
            "\n",
            "CPU times: user 299 ms, sys: 0 ns, total: 299 ms\n",
            "Wall time: 201 ms\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "result = translator.translate(\n",
        "    input_text = input_text)\n",
        "\n",
        "print(result['text'][0].numpy().decode())\n",
        "print(result['text'][1].numpy().decode())\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "n05evRexJEnp"
      },
      "outputs": [],
      "source": [
        "@tf.function(input_signature=[tf.TensorSpec(dtype=tf.string, shape=[None])])\n",
        "def tf_translate(self, input_text):\n",
        "    return self.translate(input_text)\n",
        "\n",
        "Translator.tf_translate = tf_translate"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "t_PXzl6YJTTA",
        "outputId": "f6ed4ebc-6780-4ed0-afe5-e6c4ecd2fbb7"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "CPU times: user 2.01 s, sys: 7.79 ms, total: 2.01 s\n",
            "Wall time: 1.88 s\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "result = translator.tf_translate(\n",
        "    input_text = input_text)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xT8bXsnQJVn1",
        "outputId": "d11cfa45-5807-45f4-c42e-e3d95539c260"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "will be convenient tomorrow on .\n",
            "were strong .\n",
            "\n",
            "CPU times: user 167 ms, sys: 12.3 ms, total: 179 ms\n",
            "Wall time: 56.8 ms\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "result = translator.tf_translate(\n",
        "    input_text = input_text)\n",
        "\n",
        "print(result['text'][0].numpy().decode())\n",
        "print(result['text'][1].numpy().decode())\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "IDlBgtcjJXZt",
        "outputId": "ee320496-e049-42b7-a3a7-f5772db9e87a"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[0.99999994 1.         1.         1.         1.         1.0000001\n",
            " 1.        ]\n"
          ]
        }
      ],
      "source": [
        "a = result['attention'][0]\n",
        "\n",
        "print(np.sum(a, axis=-1))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "PRNhMCAxJXdM",
        "outputId": "b9273a26-f7c9-4184-d9a2-64f41b83747f"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "_ = plt.bar(range(len(a[0, :])), a[0, :])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "lsPkRNkyJXn_",
        "outputId": "2f29610b-907f-4e21-81a8-5e14cb1787db"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7f1240c516f0>"
            ]
          },
          "execution_count": 88,
          "metadata": {},
          "output_type": "execute_result"
        },
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAGdCAYAAAAv9mXmAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAA9hAAAPYQGoP6dpAAAXwklEQVR4nO3df2zV9b3H8ddpDz38aisFCu3lgI2i/Cgga5kWcFNxTRolmmVOF2TNfvzRWX7ZmCn6h+4Xx/2xRXeZzcoWHFmw3N0NZbkDrJkUDeuEYiNBgiAkVKA2cPGc0m0HaL/3D68nqwjyPf2+++k5ez6Sk3hOviff10mwT77nlDbkeZ4nAAACluN6AAAgOxEYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBgIjzUJ+zv79epU6eUn5+vUCg01KcHAAyC53nq6elRaWmpcnKufo0y5IE5deqUotHoUJ8WABCgzs5OTZky5arHDHlg8vPzJUm3LXpC4XBkqE9v5h8TRrieELi/T86ud1D/47+Pu54QOO/CBdcTgpeF72x4Fy+6nhCYS95F7T7/X6mv5Vcz5IH55G2xcDiicHjkUJ/eTHhE9gUmN5JdgQnn5LmeEDgv+74WSznZ96K8LIzmtXzEkV1fQQAAwwaBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAE2kF5oUXXlBZWZlGjhypiooKvfHGG0HvAgBkON+B2bJli9asWaOnnnpKb7/9tm6//XbV1NToxIkTFvsAABnKd2B+/vOf6zvf+Y6++93vaubMmXruuecUjUbV2NhosQ8AkKF8BebChQtqb29XdXX1gMerq6u1Z8+ez3xOMplUIpEYcAMAZD9fgTlz5oz6+vo0adKkAY9PmjRJXV1dn/mcWCymwsLC1C0ajaa/FgCQMdL6kD8UCg2473neZY99Yu3atYrH46lbZ2dnOqcEAGSYsJ+DJ0yYoNzc3MuuVrq7uy+7qvlEJBJRJBJJfyEAICP5uoLJy8tTRUWFWlpaBjze0tKihQsXBjoMAJDZfF3BSFJDQ4OWL1+uyspKVVVVqampSSdOnFBdXZ3FPgBAhvIdmAcffFBnz57VD3/4Q50+fVrl5eX685//rGnTplnsAwBkKN+BkaRHHnlEjzzySNBbAABZhJ9FBgAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMBE2NWJcy72K8frd3X6wF23v9v1hMCdmznZ9YRAXSzLrtcjSeGDx11PMJDrekDgQqNHuZ4QmFB/rtRzbcdyBQMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGDCd2B2796tpUuXqrS0VKFQSC+//LLBLABApvMdmN7eXs2bN0/r16+32AMAyBJhv0+oqalRTU2NxRYAQBbxHRi/ksmkkslk6n4ikbA+JQBgGDD/kD8Wi6mwsDB1i0aj1qcEAAwD5oFZu3at4vF46tbZ2Wl9SgDAMGD+FlkkElEkErE+DQBgmOHfwQAATPi+gjl//ryOHj2aun/8+HF1dHSoqKhIU6dODXQcACBz+Q7Mvn37dOedd6buNzQ0SJJqa2v14osvBjYMAJDZfAfmjjvukOd5FlsAAFmEz2AAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmAi7OvHFsWF5I5ydPnDh/R+6nhC4spey6+8fhx4tcj0hcLOeGeV6QuC8v//D9YTAeecvuJ4QGM+79teSXV9BAADDBoEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABMEBgAgAlfgYnFYlqwYIHy8/NVXFys+++/X4cPH7baBgDIYL4C09raqvr6erW1tamlpUWXLl1SdXW1ent7rfYBADJU2M/BO3bsGHB/48aNKi4uVnt7u770pS8FOgwAkNl8BebT4vG4JKmoqOiKxySTSSWTydT9RCIxmFMCADJE2h/ye56nhoYGLV68WOXl5Vc8LhaLqbCwMHWLRqPpnhIAkEHSDsyKFSv0zjvv6KWXXrrqcWvXrlU8Hk/dOjs70z0lACCDpPUW2cqVK7Vt2zbt3r1bU6ZMueqxkUhEkUgkrXEAgMzlKzCe52nlypXaunWrdu3apbKyMqtdAIAM5ysw9fX12rx5s1555RXl5+erq6tLklRYWKhRo0aZDAQAZCZfn8E0NjYqHo/rjjvuUElJSeq2ZcsWq30AgAzl+y0yAACuBT+LDABggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATPj6lclBGv3+/yqcG3F1+sD1X7joekLgcuI9ricE6qYXR7ueELhDT13vekLgZjSecz0hcKHus64nBCbUnyP1XtuxXMEAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCY8BWYxsZGzZ07VwUFBSooKFBVVZW2b99utQ0AkMF8BWbKlCl69tlntW/fPu3bt0933XWX7rvvPh08eNBqHwAgQ4X9HLx06dIB93/yk5+osbFRbW1tmj17dqDDAACZzVdg/lVfX59+//vfq7e3V1VVVVc8LplMKplMpu4nEol0TwkAyCC+P+Q/cOCAxo4dq0gkorq6Om3dulWzZs264vGxWEyFhYWpWzQaHdRgAEBm8B2Ym2++WR0dHWpra9P3vvc91dbW6t13373i8WvXrlU8Hk/dOjs7BzUYAJAZfL9FlpeXpxtvvFGSVFlZqb179+r555/Xr371q888PhKJKBKJDG4lACDjDPrfwXieN+AzFgAAJJ9XME8++aRqamoUjUbV09Oj5uZm7dq1Szt27LDaBwDIUL4C8+GHH2r58uU6ffq0CgsLNXfuXO3YsUNf+cpXrPYBADKUr8D85je/sdoBAMgy/CwyAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABMEBgAgAkCAwAwQWAAACbCrk7sdXXLC+W5On3gQrlZ2OpRI10vCFTo7cOuJwRu5smJricE7oP/zHc9IXCT15W6nhCYvkv/lM5e27FZ+FURADAcEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmBhUYGKxmEKhkNasWRPQHABAtkg7MHv37lVTU5Pmzp0b5B4AQJZIKzDnz5/XsmXLtGHDBo0bNy7oTQCALJBWYOrr63XPPffo7rvv/txjk8mkEonEgBsAIPuF/T6hublZ+/fv1969e6/p+Fgsph/84Ae+hwEAMpuvK5jOzk6tXr1av/vd7zRy5Mhres7atWsVj8dTt87OzrSGAgAyi68rmPb2dnV3d6uioiL1WF9fn3bv3q3169crmUwqNzd3wHMikYgikUgwawEAGcNXYJYsWaIDBw4MeOxb3/qWZsyYoccff/yyuAAA/n35Ckx+fr7Ky8sHPDZmzBiNHz/+sscBAP/e+Jf8AAATvr+L7NN27doVwAwAQLbhCgYAYILAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACAibCrE4fy8hQK5bk6ffAmFrleELj4nAmuJwSq8FKf6wmB8wrGuJ4QuPHrR7ueELgLRdnzd/lLF6/92Ox51QCAYYXAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMCEr8A888wzCoVCA26TJ0+22gYAyGBhv0+YPXu2XnvttdT93NzcQAcBALKD78CEw2GuWgAAn8v3ZzBHjhxRaWmpysrK9NBDD+nYsWNXPT6ZTCqRSAy4AQCyn6/A3Hrrrdq0aZN27typDRs2qKurSwsXLtTZs2ev+JxYLKbCwsLULRqNDno0AGD4C3me56X75N7eXt1www36/ve/r4aGhs88JplMKplMpu4nEglFo1EtGVercCgv3VMPPxOLXC8IXGLOBNcTAlX41knXEwLnFYxxPSFw/ywZ63pC4Lxw9nzD7qWL/9Se155WPB5XQUHBVY/1/RnMvxozZozmzJmjI0eOXPGYSCSiSCQymNMAADLQoLKaTCZ16NAhlZSUBLUHAJAlfAXmscceU2trq44fP66//e1v+trXvqZEIqHa2lqrfQCADOXrLbIPPvhA3/jGN3TmzBlNnDhRt912m9ra2jRt2jSrfQCADOUrMM3NzVY7AABZJnu+tQEAMKwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMBF2deK+j+IKhUa4On3gQj09ricEbsyRY64nBOqS57meELxQyPWCwEWOj3I9IXDbj+5xPSEwiZ5+jbvp2o7lCgYAYILAAABMEBgAgAkCAwAwQWAAACYIDADABIEBAJggMAAAEwQGAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMCE78CcPHlSDz/8sMaPH6/Ro0frlltuUXt7u8U2AEAGC/s5+Ny5c1q0aJHuvPNObd++XcXFxXr//fd13XXXGc0DAGQqX4H56U9/qmg0qo0bN6Yeu/7664PeBADIAr7eItu2bZsqKyv1wAMPqLi4WPPnz9eGDRuu+pxkMqlEIjHgBgDIfr4Cc+zYMTU2Nmr69OnauXOn6urqtGrVKm3atOmKz4nFYiosLEzdotHooEcDAIa/kOd53rUenJeXp8rKSu3Zsyf12KpVq7R371799a9//cznJJNJJZPJ1P1EIqFoNKo7QvcrHBoxiOnDSyg31/WEwHl9fa4nBOva/6hnjlDI9YLA5Ywa5XpC4LYf3fP5B2WIRE+/xt10TPF4XAUFBVc91tcVTElJiWbNmjXgsZkzZ+rEiRNXfE4kElFBQcGAGwAg+/kKzKJFi3T48OEBj7333nuaNm1aoKMAAJnPV2AeffRRtbW1ad26dTp69Kg2b96spqYm1dfXW+0DAGQoX4FZsGCBtm7dqpdeeknl5eX60Y9+pOeee07Lli2z2gcAyFC+/h2MJN1777269957LbYAALIIP4sMAGCCwAAATBAYAIAJAgMAMEFgAAAmCAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAATBAYAYILAAABM+P6VyYPleZ4k6ZJ3cahPbSrk9bueEDjP63M9IVj//2cvu4RcDwhcjpfrekLgEj3Z8/Uhcf7j1+Jdw/9PQx6Ynp4eSdKb+h8pm/5/v+R6AP4tZdP/Q5/4u+sBwRt3k+sFwevp6VFhYeFVjwl515KhAPX39+vUqVPKz89XKGT3t69EIqFoNKrOzk4VFBSYnWco8ZqGv2x7PRKvKVMM1WvyPE89PT0qLS1VTs7VP2UZ8iuYnJwcTZkyZcjOV1BQkDV/gD7Baxr+su31SLymTDEUr+nzrlw+wYf8AAATBAYAYCJrAxOJRPT0008rEom4nhIYXtPwl22vR+I1ZYrh+JqG/EN+AMC/h6y9ggEAuEVgAAAmCAwAwASBAQCYyMrAvPDCCyorK9PIkSNVUVGhN954w/WkQdm9e7eWLl2q0tJShUIhvfzyy64nDUosFtOCBQuUn5+v4uJi3X///Tp8+LDrWYPS2NiouXPnpv6RW1VVlbZv3+56VmBisZhCoZDWrFnjesqgPPPMMwqFQgNukydPdj1rUE6ePKmHH35Y48eP1+jRo3XLLbeovb3d9SxJWRiYLVu2aM2aNXrqqaf09ttv6/bbb1dNTY1OnDjhelraent7NW/ePK1fv971lEC0traqvr5ebW1tamlp0aVLl1RdXa3e3l7X09I2ZcoUPfvss9q3b5/27dunu+66S/fdd58OHjzoetqg7d27V01NTZo7d67rKYGYPXu2Tp8+nbodOHDA9aS0nTt3TosWLdKIESO0fft2vfvuu/rZz36m6667zvW0j3lZ5otf/KJXV1c34LEZM2Z4TzzxhKNFwZLkbd261fWMQHV3d3uSvNbWVtdTAjVu3Djv17/+tesZg9LT0+NNnz7da2lp8b785S97q1evdj1pUJ5++mlv3rx5rmcE5vHHH/cWL17sesYVZdUVzIULF9Te3q7q6uoBj1dXV2vPnj2OVuHzxONxSVJRUZHjJcHo6+tTc3Ozent7VVVV5XrOoNTX1+uee+7R3Xff7XpKYI4cOaLS0lKVlZXpoYce0rFjx1xPStu2bdtUWVmpBx54QMXFxZo/f742bNjgelZKVgXmzJkz6uvr06RJkwY8PmnSJHV1dTlahavxPE8NDQ1avHixysvLXc8ZlAMHDmjs2LGKRCKqq6vT1q1bNWvWLNez0tbc3Kz9+/crFou5nhKYW2+9VZs2bdLOnTu1YcMGdXV1aeHChTp79qzraWk5duyYGhsbNX36dO3cuVN1dXVatWqVNm3a5HqaJAc/TXkofPrXAHieZ/qrAZC+FStW6J133tGbb77pesqg3Xzzzero6NBHH32kP/zhD6qtrVVra2tGRqazs1OrV6/Wq6++qpEjR7qeE5iamprUf8+ZM0dVVVW64YYb9Nvf/lYNDQ0Ol6Wnv79flZWVWrdunSRp/vz5OnjwoBobG/XNb37T8bosu4KZMGGCcnNzL7ta6e7uvuyqBu6tXLlS27Zt0+uvvz6kv8LBSl5enm688UZVVlYqFotp3rx5ev75513PSkt7e7u6u7tVUVGhcDiscDis1tZW/eIXv1A4HFZfX3b8ttMxY8Zozpw5OnLkiOspaSkpKbnsLzAzZ84cNt/UlFWBycvLU0VFhVpaWgY83tLSooULFzpahU/zPE8rVqzQH//4R/3lL39RWVmZ60kmPM9TMpl0PSMtS5Ys0YEDB9TR0ZG6VVZWatmyZero6FBubnb8WuNkMqlDhw6ppKTE9ZS0LFq06LJv8X/vvfc0bdo0R4sGyrq3yBoaGrR8+XJVVlaqqqpKTU1NOnHihOrq6lxPS9v58+d19OjR1P3jx4+ro6NDRUVFmjp1qsNl6amvr9fmzZv1yiuvKD8/P3XFWVhYqFGjRjlel54nn3xSNTU1ikaj6unpUXNzs3bt2qUdO3a4npaW/Pz8yz4TGzNmjMaPH5/Rn5U99thjWrp0qaZOnaru7m79+Mc/ViKRUG1tretpaXn00Ue1cOFCrVu3Tl//+tf11ltvqampSU1NTa6nfcztN7HZ+OUvf+lNmzbNy8vL877whS9k/Le/vv76654+/u3rA261tbWup6Xls16LJG/jxo2up6Xt29/+durP3MSJE70lS5Z4r776qutZgcqGb1N+8MEHvZKSEm/EiBFeaWmp99WvftU7ePCg61mD8qc//ckrLy/3IpGIN2PGDK+pqcn1pBR+XD8AwERWfQYDABg+CAwAwASBAQCYIDAAABMEBgBggsAAAEwQGACACQIDADBBYAAAJggMAMAEgQEAmCAwAAAT/wcAgWHj/s2DAQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.imshow(np.array(a), vmin=0.0)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "u528ctdll6pw"
      },
      "outputs": [],
      "source": [
        "def plot_attention(attention, sentence, predicted_sentence):\n",
        "    sentence = tf_lower_and_split_punct(sentence).numpy().decode().split()\n",
        "    predicted_sentence = predicted_sentence.numpy().decode().split() + ['[END]']\n",
        "    fig = plt.figure(figsize=(10, 10))\n",
        "    ax = fig.add_subplot(1, 1, 1)\n",
        "\n",
        "    attention = attention[:len(predicted_sentence), :len(sentence)]\n",
        "\n",
        "    ax.matshow(attention, cmap='viridis', vmin=0.0)\n",
        "\n",
        "    fontdict = {'fontsize': 14}\n",
        "\n",
        "    ax.set_xticklabels([''] + sentence, fontdict=fontdict, rotation=90)\n",
        "    ax.set_yticklabels([''] + predicted_sentence, fontdict=fontdict)\n",
        "\n",
        "    ax.xaxis.set_major_locator(ticker.MultipleLocator(1))\n",
        "    ax.yaxis.set_major_locator(ticker.MultipleLocator(1))\n",
        "\n",
        "    ax.set_xlabel('Input text')\n",
        "    ax.set_ylabel('Output text')\n",
        "    plt.suptitle('Attention weights')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "DYIIpTRTJXtF",
        "outputId": "3b033e4a-d0e2-49c3-8640-ecc3bfbc18bc"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/tmp/wsuser/ipykernel_416/3262901620.py:13: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_xticklabels([''] + sentence, fontdict=fontdict, rotation=90)\n",
            "/tmp/wsuser/ipykernel_416/3262901620.py:14: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_yticklabels([''] + predicted_sentence, fontdict=fontdict)\n"
          ]
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x1000 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "i=0\n",
        "plot_attention(result['attention'][i], input_text[i], result['text'][i])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "oHpXhKMnJXvQ",
        "outputId": "db0be390-06da-4fe0-f0bf-c3d2eb5b235a"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "please sing .\n",
            "will you be able to walk ?\n",
            "speak to me .\n",
            "\n",
            "CPU times: user 192 ms, sys: 12.2 ms, total: 204 ms\n",
            "Wall time: 66 ms\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "three_input_text = tf.constant([\n",
        "    # This is my life.\n",
        "    \"chante s'il te plait.\",\n",
        "    # Are they still home?\n",
        "    'devrait-on marcher?',\n",
        "    # Try to find out.'\n",
        "    'soit content.',\n",
        "])\n",
        "\n",
        "result = translator.tf_translate(three_input_text)\n",
        "\n",
        "for tr in result['text']:\n",
        "    print(tr.numpy().decode())\n",
        "\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "imk9VI2HJXzB",
        "outputId": "48815dd8-2755-4432-aff8-ace7d5e22ff3"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<tf.Tensor: shape=(3,), dtype=string, numpy=\n",
              "array([b'please sing .', b'will you be able to walk ?', b'speak to me .'],\n",
              "      dtype=object)>"
            ]
          },
          "execution_count": 92,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "result['text']"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "JSKr6uQoJe66",
        "outputId": "1dc2caa9-d4b4-4902-b94e-1b4f574edebf"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/tmp/wsuser/ipykernel_416/3262901620.py:13: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_xticklabels([''] + sentence, fontdict=fontdict, rotation=90)\n",
            "/tmp/wsuser/ipykernel_416/3262901620.py:14: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_yticklabels([''] + predicted_sentence, fontdict=fontdict)\n"
          ]
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x1000 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "i = 0\n",
        "plot_attention(result['attention'][i], three_input_text[i], result['text'][i])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ckyjNqWJJfmg",
        "outputId": "4f6c7d4a-0c60-42ba-ab92-d06d565a7de3"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/tmp/wsuser/ipykernel_416/3262901620.py:13: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_xticklabels([''] + sentence, fontdict=fontdict, rotation=90)\n",
            "/tmp/wsuser/ipykernel_416/3262901620.py:14: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_yticklabels([''] + predicted_sentence, fontdict=fontdict)\n"
          ]
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x1000 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "i = 1\n",
        "plot_attention(result['attention'][i], three_input_text[i], result['text'][i])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Tri8JtgsJhaF",
        "outputId": "e0d65557-0839-4338-f28d-58b4c3cd2172"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/tmp/wsuser/ipykernel_416/3262901620.py:13: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_xticklabels([''] + sentence, fontdict=fontdict, rotation=90)\n",
            "/tmp/wsuser/ipykernel_416/3262901620.py:14: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_yticklabels([''] + predicted_sentence, fontdict=fontdict)\n"
          ]
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x1000 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "i = 2\n",
        "plot_attention(result['attention'][i], three_input_text[i], result['text'][i])"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "BiTnINVSJiSa",
        "outputId": "64591e13-47be-4ee3-d05f-dd84aee9273f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Expected output:\n",
            " It may be impossible to get a completely error-free corpus due to the\n",
            "nature of this kind of collaborative effort. However, if we encourage\n",
            "members to contribute sentences in their own languages rather than\n",
            "experiment in languages they are learning, we might be able to\n",
            "minimize errors.\n"
          ]
        }
      ],
      "source": [
        "long_input_text = tf.constant([inp[-1]])\n",
        "\n",
        "import textwrap\n",
        "print('Expected output:\\n', '\\n'.join(textwrap.wrap(targ[-1])))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "KM1mIP0uJkNr",
        "outputId": "f7e98946-368c-4fa7-91d0-325ba6462a64"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/tmp/wsuser/ipykernel_416/3262901620.py:13: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_xticklabels([''] + sentence, fontdict=fontdict, rotation=90)\n",
            "/tmp/wsuser/ipykernel_416/3262901620.py:14: UserWarning: FixedFormatter should only be used together with FixedLocator\n",
            "  ax.set_yticklabels([''] + predicted_sentence, fontdict=fontdict)\n"
          ]
        },
        {
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1000x1000 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "result = translator.tf_translate(long_input_text)\n",
        "\n",
        "i = 0\n",
        "plot_attention(result['attention'][i], long_input_text[i], result['text'][i])\n",
        "_ = plt.suptitle('This never works')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "fuJGqjujJmPH",
        "outputId": "7be54fa0-2933-4e06-e692-a1c8e59d69e1"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "WARNING:absl:Found untraced functions such as encoder_2_layer_call_fn, encoder_2_layer_call_and_return_conditional_losses, decoder_2_layer_call_fn, decoder_2_layer_call_and_return_conditional_losses, embedding_4_layer_call_fn while saving (showing 5 of 24). These functions will not be directly callable after loading.\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Assets written to: translator/assets\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Assets written to: translator/assets\n"
          ]
        }
      ],
      "source": [
        "# sauvegarde du model\n",
        "\"\"\"\n",
        "tf.saved_model.save(translator, 'translator',\n",
        "                    signatures={'serving_default': translator.tf_translate})\n",
        "\"\"\"\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "FDZCcgDPl6py"
      },
      "source": [
        "## Nous importons le model déja entrainé"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "joW26D35l6pz"
      },
      "outputs": [],
      "source": [
        "\n",
        "# @hidden_cell\n",
        "# The following code contains the credentials for a file in your IBM Cloud Object Storage.\n",
        "# You might want to remove those credentials before you share your notebook.\n",
        "metadata_5 = {\n",
        "    'IAM_SERVICE_ID': 'iam-ServiceId-1-450d-9aef-59f60ec1aab9',\n",
        "    'IBM_API_KEY_ID': '_K9kKnfQ26uY7SJkmoIEEOCyJYOB',\n",
        "    'ENDPOINT': 'https://s3.private.eu.cloud-object-storage.appdomain.cloud',\n",
        "    'IBM_AUTH_ENDPOINT': 'https://iam.cloud.ibm.com/oidc/token',\n",
        "    'BUCKET': 'nlptranslationseq2seqwithattentio-donotdelete-pr-medpzvaa1j3gyb',\n",
        "    'FILE': 'translator_Tarek_corpus_160K_folder.zip'\n",
        "}\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3TCjk4YHl6p0"
      },
      "outputs": [],
      "source": [
        "from ibm_botocore.client import Config\n",
        "import ibm_boto3\n",
        "def download_file_cos(credentials,local_file_name,key):  \n",
        "    cos = ibm_boto3.client(service_name='s3',\n",
        "    ibm_api_key_id=credentials['IBM_API_KEY_ID'],\n",
        "    ibm_service_instance_id=credentials['IAM_SERVICE_ID'],\n",
        "    ibm_auth_endpoint=credentials['IBM_AUTH_ENDPOINT'],\n",
        "    config=Config(signature_version='oauth'),\n",
        "    endpoint_url=credentials['ENDPOINT'])\n",
        "    try:\n",
        "        res=cos.download_file(Bucket=credentials['BUCKET'],Key=key,Filename=local_file_name)\n",
        "    except Exception as e:\n",
        "        print(Exception, e)\n",
        "    else:\n",
        "        print('File Downloaded')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "eg8w0txal6p0",
        "outputId": "2c5a2fe7-a274-431d-d2cd-49f1cc5d7a43"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "File Downloaded\n"
          ]
        }
      ],
      "source": [
        "download_file_cos(metadata_5,'translator_Tarek_corpus_160K_folder.zip','translator_Tarek_corpus_160K_folder.zip')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "P2PklRWCl6p0"
      },
      "outputs": [],
      "source": [
        "from zipfile import ZipFile \n",
        "with ZipFile ('translator_Tarek_corpus_160K_folder.zip', 'r') as zipObj : \n",
        "    #Extraire tt les fichiers dans le directory : \n",
        "    zipObj.extractall()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "yzZQiLbdl6p0",
        "outputId": "0810fde9-9013-4997-f1d7-d6c8b1c03894"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['variables',\n",
              " 'assets',\n",
              " 'saved_model.pb',\n",
              " '__MACOSX',\n",
              " 'translator_Tarek_corpus_160K.zip',\n",
              " '.virtual_documents',\n",
              " 'translator_Tarek_corpus_160K_folder.zip',\n",
              " 'translator_Tarek_corpus_160K_folder']"
            ]
          },
          "execution_count": 65,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "import os\n",
        "os.listdir()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "cb4gRwAll6p1"
      },
      "outputs": [],
      "source": [
        "os.chdir(\"translator_Tarek_corpus_160K_folder\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "pVh-2wb1Jm8i"
      },
      "outputs": [],
      "source": [
        "reloaded = tf.saved_model.load('translator')\n",
        "three_input_text = tf.constant([\n",
        "    # This is my life.\n",
        "    \"chante s'il te plait.\",\n",
        "    # Are they still home?\n",
        "    'devrait-on marcher?',\n",
        "    # Try to find out.'\n",
        "    'soit content.',\n",
        "])\n",
        "\n",
        "result = reloaded.tf_translate(three_input_text)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "8Rl9PcXTJoS6",
        "outputId": "1d4b72a7-2876-43c5-e406-c9456da59f17"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "please sing .\n",
            "do you use walking ?\n",
            "is nothing glad .\n",
            "\n",
            "CPU times: user 147 ms, sys: 17.2 ms, total: 164 ms\n",
            "Wall time: 53.9 ms\n"
          ]
        }
      ],
      "source": [
        "%%time\n",
        "result = reloaded.tf_translate(three_input_text)\n",
        "\n",
        "for tr in result['text']:\n",
        "  print(tr.numpy().decode())\n",
        "\n",
        "print()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "xOljwVaEl6p1"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "uBEQIrZEl6p2"
      },
      "outputs": [],
      "source": []
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Ul9de0RGl6p2"
      },
      "outputs": [],
      "source": []
    }
  ],
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3.10",
      "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": "3.10.6"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}