Lati.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Db4mybGWBpem"
      },
      "outputs": [],
      "source": [
        "!pip install stopwords\n",
        "\n",
        "import stopwords\n",
        "\n",
        "import tensorflow as tf\n",
        "\n",
        "import tensorflow_datasets as tfds\n",
        "\n",
        "tfds.disable_progress_bar()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train, test = tfds.load(\"imdb_reviews\", split=[\"train\", \"test\"], as_supervised=True)"
      ],
      "metadata": {
        "id": "57xZ6MccCP6B"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "unlist = stopwords.get_stopwords(\"english\")\n",
        "\n",
        "type(unlist)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WfXgIJjKCUyA",
        "outputId": "cc4acec4-dade-4b4f-e576-e6a5d1472ca2"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "list"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_input = []; train_label = []\n",
        "\n",
        "test_input = []; test_label = []\n",
        "\n",
        "retori = lambda snt : [i for i in snt.lower().split() if i not in unlist]\n",
        "\n",
        "for data, label in train:\n",
        "  retari = \" \".join(retori(data.numpy().decode(\"utf8\")))\n",
        "  train_input.append(retari)\n",
        "  train_label.append(label.numpy())\n",
        "\n",
        "for data, label in test:\n",
        "  retari = \" \".join(retori(data.numpy().decode(\"utf8\")))\n",
        "  test_input.append(retari)\n",
        "  test_label.append(label.numpy())"
      ],
      "metadata": {
        "id": "qRnpWcvkDI7d"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "len(train_input), len(test_input), len(train_label), len(test_label)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "A9UVOml1DLlq",
        "outputId": "55151279-9d22-431e-a16a-9eb13cb9792e"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000, 25000, 25000, 25000)"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import tensorflow.keras.layers as layers\n",
        "\n",
        "vector = layers.TextVectorization(max_tokens=10)\n",
        "\n",
        "sample = [\"aku suka ikan\", \"ikan goreng\", \"ibu masak ikan\", \"bapak masak ayam bersama ibu\"]\n",
        "\n",
        "vector.adapt(sample)"
      ],
      "metadata": {
        "id": "nO54BYqzDk3H"
      },
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "vector.get_vocabulary()[:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9lqG1_NiENHl",
        "outputId": "75c7130c-e613-4e99-ce3e-ec0cde7134ec"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['', '[UNK]', 'ikan', 'masak', 'ibu']"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "vector.get_vocabulary()[5:]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "XJi2sUaGEgUF",
        "outputId": "35e6f7f4-5fc1-433c-8981-8333afbf7ef5"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['suka', 'goreng', 'bersama', 'bapak', 'ayam']"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "vector.get_vocabulary()[1]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "EvBYz_ZzE4tQ",
        "outputId": "6a44b3c0-44c7-4ca5-b48d-944aee2ce8f6"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'[UNK]'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "samplet = [\"ikan masak ibu tadi\"]\n",
        "\n",
        "vector(samplet).numpy()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ef-chS6nE6x_",
        "outputId": "ce7e61fa-7db2-4258-9b8e-8121263c79d2"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[2, 3, 4, 1]])"
            ]
          },
          "metadata": {},
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "vector.get_vocabulary(include_special_tokens=False)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Mxc5zkvYE9XV",
        "outputId": "0223209a-3722-4996-ca34-bee95cb74d5f"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['ikan', 'masak', 'ibu', 'suka', 'goreng', 'bersama', 'bapak', 'ayam']"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "len(vector.get_vocabulary())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nZCVDDBgE_H5",
        "outputId": "55ff4cd7-320d-41a9-a0f2-6498b136e93f"
      },
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "10"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "max_tokens = 15000\n",
        "\n",
        "embedding = 16\n",
        "\n",
        "output_sequence_length = 180\n",
        "\n",
        "tokener = layers.TextVectorization(\n",
        "  max_tokens=max_tokens, output_mode=\"int\",\n",
        "  pad_to_max_tokens=True,\n",
        "  output_sequence_length=output_sequence_length)\n",
        "\n",
        "tokener.adapt(train_input)"
      ],
      "metadata": {
        "id": "K_OMc31OFBFI"
      },
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "len(tokener.get_vocabulary())"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1G2EgdokFFiA",
        "outputId": "5d8bfbc7-1b59-46cb-d43a-c85e053011cc"
      },
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "15000"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "\n",
        "train_input = np.array(tokener(train_input))\n",
        "\n",
        "train_label = np.array(train_label)"
      ],
      "metadata": {
        "id": "uTsPLA7XFHzT"
      },
      "execution_count": 15,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "len(train_input), len(train_label)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "d2HIbSXqFJ4O",
        "outputId": "d0111432-a513-4fb5-e0ac-8eb3f64edd7a"
      },
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(25000, 25000)"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_input[0][:5]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uyOLVJ6VFNhV",
        "outputId": "f8634181-430f-48f1-b201-07f69fe053ee"
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([  318,   292,     3, 10671,  1244])"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_input = np.array(tokener(test_input))\n",
        "\n",
        "test_label = np.array(test_label)"
      ],
      "metadata": {
        "id": "Cck5XLHiFPEP"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "inputs = tf.keras.layers.Input(shape=(output_sequence_length,))\n",
        "\n",
        "embedding = tf.keras.layers.Embedding(max_tokens,embedding)(inputs)\n",
        "\n",
        "droupout = tf.keras.layers.Dropout(0.2)(embedding)\n",
        "\n",
        "pooling = tf.keras.layers.GlobalAveragePooling1D()(droupout)\n",
        "\n",
        "droupout = tf.keras.layers.Dropout(0.2)(pooling)\n",
        "\n",
        "output = tf.keras.layers.Dense(1, activation='sigmoid')(droupout)\n",
        "\n",
        "model = tf.keras.Model(inputs=inputs, outputs=output)"
      ],
      "metadata": {
        "id": "8DyStBadFRt7"
      },
      "execution_count": 19,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(optimizer=\"adam\", loss=\"binary_crossentropy\", metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "kwWd3b4lFW4Q"
      },
      "execution_count": 20,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plot = model.fit(train_input, train_label, epochs=10, validation_data=(test_input, test_label))"
      ],
      "metadata": {
        "id": "UokyPwCG9Nvk"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "model.evaluate(train_input, train_label, verbose=0)"
      ],
      "metadata": {
        "id": "Q-St6Wq79pWe",
        "outputId": "88ee018e-8159-4592-8381-0db56d9aefe0",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.1617421805858612, 0.9460399746894836]"
            ]
          },
          "metadata": {},
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.evaluate(test_input, test_label, verbose=0)"
      ],
      "metadata": {
        "id": "5UOgiZ2U9rAD",
        "outputId": "fdf7139a-e598-4621-ac33-3cb402f256ff",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.3069874346256256, 0.8759599924087524]"
            ]
          },
          "metadata": {},
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import Helplot, os\n",
        "\n",
        "plot = Helplot.Helplot(plot)"
      ],
      "metadata": {
        "id": "9UCjVqR79tQ_"
      },
      "execution_count": 24,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "plot.trainplot"
      ],
      "metadata": {
        "id": "MPuPKAbM9yRP",
        "outputId": "dffcb319-9188-4145-92e4-0ad696960c12",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 512
        }
      },
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x500 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot.testiplot"
      ],
      "metadata": {
        "id": "avPMytcF9ytn",
        "outputId": "f7e2daf3-65dd-4213-f9d4-371f47aff9b7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 512
        }
      },
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1000x500 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import os\n",
        "\n",
        "model.save(\"/content/model.h5\")"
      ],
      "metadata": {
        "id": "OfDE8mxc91vK"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!pip install apache_beam[gcp] --quiet\n",
        "\n",
        "import warnings\n",
        "\n",
        "import apache_beam as beam\n",
        "\n",
        "warnings.filterwarnings(\"ignore\")"
      ],
      "metadata": {
        "id": "6T2estcZ93uM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from apache_beam.ml.inference.tensorflow_inference import TFModelHandlerTensor\n",
        "\n",
        "from apache_beam.ml.inference.base import RunInference\n",
        "\n",
        "tensor = TFModelHandlerTensor(\"/content/model.h5\", compile=False)"
      ],
      "metadata": {
        "id": "bllR3ZWg-jmi"
      },
      "execution_count": 29,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "trainer = train_input[:10]\n",
        "\n",
        "len(trainer)"
      ],
      "metadata": {
        "id": "uVKMNUF8-mSM",
        "outputId": "5556b77b-100b-4ec7-da43-c20d00412065",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "10"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "labeler = lambda result : result.inference.numpy()[0].round()"
      ],
      "metadata": {
        "id": "B6JwwWIq-orz"
      },
      "execution_count": 31,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "with beam.Pipeline() as pipeline:\n",
        "  runner = (\n",
        "    pipeline\n",
        "    | 'Create' >> beam.Create(trainer)\n",
        "    | 'RunInference' >> RunInference(tensor)\n",
        "    | 'Outer' >> beam.Map(labeler)\n",
        "    | 'Unlist' >> beam.combiners.ToList()\n",
        "    | 'Output' >> beam.io.WriteToText(\"/content/result.txt\")\n",
        "  )"
      ],
      "metadata": {
        "id": "GdwJVw1xQFdU"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!cat /content/result.txt-00000-of-00001"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BHwKcpH8TYoj",
        "outputId": "33a1907b-eb7c-4238-b777-5878ea13f6ac"
      },
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[0.0, 0.0, 1.0, 1.0, 1.0, 1.0, 0.0, 1.0, 0.0, 0.0]\n"
          ]
        }
      ]
    }
  ]
}