Lab 01.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "source": [
        "## Lab 01"
      ],
      "metadata": {
        "id": "iyT7D5PmCE5b"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "Vj2UvzGHB0S2"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "\n",
        "import os, pathlib, warnings, urllib, shutil\n",
        "\n",
        "warnings.filterwarnings(\"ignore\")"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "giturl = \"/~https://github.com/dicodingacademy/assets/releases/download/release/\"\n",
        "\n",
        "dataset = \"rockpaperscissors.zip\"\n",
        "\n",
        "urllib.request.urlretrieve(os.path.join(giturl, dataset), dataset)\n",
        "\n",
        "shutil.unpack_archive('rockpaperscissors.zip', '/content', 'zip')"
      ],
      "metadata": {
        "id": "T1cU6OFvCZGM"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "root = \"/content/rockpaperscissors/rps-cv-images\"\n",
        "\n",
        "os.listdir(root)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fuBn5Q1-CtBf",
        "outputId": "90ae707d-57e3-459c-88c5-bfd4de5ac819"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['rock', 'paper', 'scissors', 'README_rpc-cv-images.txt']"
            ]
          },
          "metadata": {},
          "execution_count": 3
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator\n",
        "\n",
        "ImageDataGenerator = ImageDataGenerator(\n",
        "  rescale = 1.0 / 255.0,\n",
        "  rotation_range = 20,\n",
        "  zoom_range = 0.2,\n",
        "  shear_range = 0.2,\n",
        "  fill_mode = \"nearest\",\n",
        "  validation_split = 0.4\n",
        ")\n",
        "\n",
        "train = ImageDataGenerator.flow_from_directory(\n",
        "  root,\n",
        "  target_size = (150, 150),\n",
        "  batch_size = 128,\n",
        "  class_mode = \"categorical\",\n",
        "  subset = \"training\"\n",
        ")\n",
        "\n",
        "validation = ImageDataGenerator.flow_from_directory(\n",
        "  root,\n",
        "  target_size = (150, 150),\n",
        "  batch_size = 128,\n",
        "  class_mode = \"categorical\",\n",
        "  subset = \"validation\"\n",
        ")"
      ],
      "metadata": {
        "id": "7HztsKoSCueJ"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "labeler = list(train.class_indices.keys())\n",
        "\n",
        "labeler"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_dC2vMMlCvwe",
        "outputId": "63e8c854-5025-4fb9-8579-4cbfe9b36719"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['paper', 'rock', 'scissors']"
            ]
          },
          "metadata": {},
          "execution_count": 5
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### `Model Spirior`"
      ],
      "metadata": {
        "id": "_Tcxqpp4Cxjc"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "\n",
        "from tensorflow.keras import Model, layers\n",
        "\n",
        "model = Sequential([\n",
        "  layers.Conv2D(16, 3, activation=\"relu\", input_shape=(150, 150, 3)),\n",
        "  layers.MaxPool2D((2, 2)),\n",
        "  layers.Conv2D(32, 3, activation=\"relu\"),\n",
        "  layers.MaxPool2D((2, 2)),\n",
        "  layers.Conv2D(64, 3, activation=\"relu\"),\n",
        "  layers.MaxPool2D((2, 2)),\n",
        "  layers.Flatten(),\n",
        "  layers.Dropout(0.3),\n",
        "  layers.Dense(128, activation=\"relu\"),\n",
        "  layers.Dense(3, activation=\"softmax\")\n",
        "])\n",
        "\n",
        "loss = \"categorical_crossentropy\"\n",
        "\n",
        "model.compile(optimizer=\"adam\", loss=loss, metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "J0Q2rF8KDIgN"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "hist = model.fit(train, epochs=10, validation_data=validation)"
      ],
      "metadata": {
        "id": "IldlupkYDMKL"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### `Model Test`"
      ],
      "metadata": {
        "id": "XRma8yJREGrM"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.evaluate(train, verbose=False)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0zMCdPd0EIf2",
        "outputId": "93b99569-add7-424e-f10d-57c241571148"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.12937608361244202, 0.9604261517524719]"
            ]
          },
          "metadata": {},
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "model.evaluate(validation, verbose=False)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zRJx6sLlEJ4V",
        "outputId": "c7a20fff-1e1b-43ed-9970-fa796efc2782"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "[0.18601451814174652, 0.9393592476844788]"
            ]
          },
          "metadata": {},
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "*Our Model Achieves an Impressive 0.94 Average Accuracy on Both Training and Test Datasets*"
      ],
      "metadata": {
        "id": "2YR-xpSUEODY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "if not os.path.exists(\"Helplot.py\"):\n",
        "\n",
        "  giturl = \"https://raw.githubusercontent.com/Kelniit/Hut/refs/heads/main/Helplot.py\"\n",
        "\n",
        "  urllib.request.urlretrieve(giturl, \"Helplot.py\")"
      ],
      "metadata": {
        "id": "WjbvMREHERN0"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from Helplot import Helplot\n",
        "\n",
        "plot = Helplot(hist)\n",
        "\n",
        "plot.Relplot"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 465
        },
        "id": "oovIAWYtEvv6",
        "outputId": "85888331-a85d-4cfa-e114-71adaf034262"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plot.Falplot"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 465
        },
        "id": "d0EoTMm8EweM",
        "outputId": "424c06d8-49e1-4622-d375-90492bf7476c"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ModelFolder = \"/content/Modelite\"\n",
        "\n",
        "model.export(ModelFolder)"
      ],
      "metadata": {
        "id": "3dvd2GBgEyzl"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### `TensorFlow Lite`"
      ],
      "metadata": {
        "id": "wgpkOgc7E10o"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "tflite = tf.lite.TFLiteConverter\n",
        "\n",
        "tflite_model = tflite.from_saved_model(ModelFolder)\n",
        "\n",
        "tflite_model.optimizations = [tf.lite.Optimize.DEFAULT]\n",
        "\n",
        "tflite_model = tflite_model.convert()\n",
        "\n",
        "pathfile = \"HutModel01.tflite\"\n",
        "\n",
        "tflite_model_file = pathlib.Path(pathfile)\n",
        "\n",
        "tflite_model_file.write_bytes(tflite_model)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "z718m6CyE1FT",
        "outputId": "00dce3ee-bf1a-4182-debb-f919385f3b30"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2401440"
            ]
          },
          "metadata": {},
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "os.path.getsize(pathfile) / float(2**20)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dQp_PCvRE7d6",
        "outputId": "d6450976-0529-4ba1-ff0a-64c456cfb86d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2.290191650390625"
            ]
          },
          "metadata": {},
          "execution_count": 15
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### `Test TensorFlow Lite Model`"
      ],
      "metadata": {
        "id": "fHxa_gK1E-b1"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "interpreter = tf.lite.Interpreter(model_path=pathfile)\n",
        "\n",
        "interpreter.get_signature_list()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "LLpfrk2NE_zK",
        "outputId": "460a788b-c7a9-492f-db03-d89d06aeb69f"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'serving_default': {'inputs': ['keras_tensor'], 'outputs': ['output_0']}}"
            ]
          },
          "metadata": {},
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "lite_interpreter = interpreter.get_signature_runner('serving_default')\n",
        "\n",
        "lite_interpreter"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4Nq3vEmVFYBA",
        "outputId": "583014ec-975a-4fdb-81de-54a83b8adfa5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.lite.python.interpreter.SignatureRunner at 0x7bbdd66fb730>"
            ]
          },
          "metadata": {},
          "execution_count": 17
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "paper = \"/content/rockpaperscissors/paper/04l5I8TqdzF9WDMJ.png\"\n",
        "\n",
        "tester = tf.keras.utils.load_img(paper, target_size=(150, 150))\n",
        "\n",
        "tester = tf.keras.utils.img_to_array(tester)\n",
        "\n",
        "tester = tf.expand_dims(tester, axis=0)\n",
        "\n",
        "tester = tf.cast(tester / 255.0, tf.float32)\n",
        "\n",
        "tester.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uvhmf9PmFZmT",
        "outputId": "e0887a40-1e0f-4267-d682-1380631c3ebb"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "TensorShape([1, 150, 150, 3])"
            ]
          },
          "metadata": {},
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "testa = lite_interpreter(keras_tensor=tester)['output_0']\n",
        "\n",
        "testa"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "xH4QmFXnFcGt",
        "outputId": "721ac6ac-3c2e-42cb-93c8-65f96151c199"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[1.0000000e+00, 7.9003415e-15, 1.9054775e-24]], dtype=float32)"
            ]
          },
          "metadata": {},
          "execution_count": 19
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "labeler[tf.math.argmax(testa[0])]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 36
        },
        "id": "wl1V4p97FdZf",
        "outputId": "b7c085c6-bb02-4d6f-bdc5-49279f409b65"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'paper'"
            ],
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            }
          },
          "metadata": {},
          "execution_count": 20
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_img, test_label = next(iter(validation))\n",
        "\n",
        "logits = model.predict(test_img, verbose=0)\n",
        "\n",
        "logits.shape"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Ywy1KxtIFfFg",
        "outputId": "1c1ec3cf-acc9-40f2-b7c1-3c5186b06834"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(128, 3)"
            ]
          },
          "metadata": {},
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "#### `TensorFlow Serving`"
      ],
      "metadata": {
        "id": "69_u8nrKFhkZ"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "!apt update\n",
        "\n",
        "!apt-get install tensorflow-model-server"
      ],
      "metadata": {
        "id": "2z6xePsCF64M"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "class Outlabel(layers.Layer):\n",
        "  def __init__(self):\n",
        "    super(Outlabel, self).__init__()\n",
        "    self.label = labeler\n",
        "\n",
        "  def call(self, inputs):\n",
        "    tf.config.run_functions_eagerly(True)\n",
        "    logits = tf.nn.softmax(inputs)\n",
        "    logits = tf.math.argmax(inputs, axis=1)\n",
        "    result = tf.gather(self.label, logits)\n",
        "    return result"
      ],
      "metadata": {
        "id": "7flf06R1GWZw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "input_layer = tf.keras.layers.Input(shape=(150, 150, 3))\n",
        "\n",
        "modular = model(input_layer)\n",
        "\n",
        "output_layer = Outlabel()(modular)\n",
        "\n",
        "modular_server = tf.keras.Model(input_layer, output_layer)\n",
        "\n",
        "modular_server.output"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "61YJ3-yVGYHs",
        "outputId": "f12781f7-ab61-48c3-fe22-0831969b3f19"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<KerasTensor shape=(None,), dtype=string, sparse=False, name=keras_tensor_13>"
            ]
          },
          "metadata": {},
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "tester = modular_server.predict(test_img, verbose=False)\n",
        "\n",
        "tester[:3]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t3p70_TNGZgL",
        "outputId": "dcdbb25a-14d1-4bf7-fca8-8604aaf8d1f6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([b'paper', b'rock', b'paper'], dtype=object)"
            ]
          },
          "metadata": {},
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "ModelFolder = \"/content/ModelServer\"\n",
        "\n",
        "ModelSerial = \"1\"\n",
        "\n",
        "ModuleFolder = os.path.join(ModelFolder, ModelSerial)\n",
        "\n",
        "modular_server.export(ModuleFolder)"
      ],
      "metadata": {
        "id": "_XTPuExsGazm"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "os.environ[\"ModelFolder\"] = ModelFolder"
      ],
      "metadata": {
        "id": "gJYH27VXGcfj"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "%%bash --bg\n",
        "nohup tensorflow_model_server \\\n",
        "  --rest_api_port=8501 \\\n",
        "  --model_name=HutModule \\\n",
        "  --model_base_path=\"${ModelFolder}\" > server.log 2>&1"
      ],
      "metadata": {
        "id": "gtjvR6TtGe7f"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "!tail server.log"
      ],
      "metadata": {
        "id": "hCjXhaEbGfaw"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import json, requests\n",
        "\n",
        "test_img = test_img.tolist()\n",
        "\n",
        "url = \"http://127.0.0.1:8501/v1/models/HutModule:predict\"\n",
        "\n",
        "image = json.dumps({\"signature_name\":\"serving_default\", \"instances\":test_img})\n",
        "\n",
        "content = {\"content-type\":\"application/json\"}\n",
        "\n",
        "response = requests.post(url, data=image, headers=content)\n",
        "\n",
        "loader = json.loads(response.text)\n",
        "\n",
        "result = loader[\"predictions\"]\n",
        "\n",
        "result[:4]"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "go1GP31nGkOm",
        "outputId": "e32630ba-9c84-4d2c-a35d-ea7e66e46df6"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['paper', 'rock', 'paper', 'scissors']"
            ]
          },
          "metadata": {},
          "execution_count": 30
        }
      ]
    }
  ]
}