MNIST

mnist.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "MNIST",
      "provenance": [],
      "collapsed_sections": [
        "HJeM_36aA8lN"
      ],
      "authorship_tag": "ABX9TyO7Odq05M3XMEOBhN5QDZPX",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "gpuClass": "standard"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/gist/mattslight/2faf5a1f62cb73f7bf14885926edd759/mnist.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# The Traditional MNIST Digital Detection ML workflow\n",
        "\n",
        "## Outline\n",
        "\n",
        "1.   Install wandb (A tool used to graph key learning metrics and optimisation)\n",
        "2.   Download the MNIST dataset\n",
        "3. Check the size and shape of the dataset\n",
        "4. Build the model\n",
        "5. Compile the model\n",
        "6. Train the model \n",
        "7. Test the the model\n"
      ],
      "metadata": {
        "id": "jZAe55h_-xYK"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 1. Install wandb"
      ],
      "metadata": {
        "id": "HJeM_36aA8lN"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "### Install and initalise wandb for graphing\n",
        "!pip install wandb &>/dev/null\n",
        "!wandb login\n",
        "import wandb\n",
        "from wandb.keras import WandbCallback\n",
        "wandb.init()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 180
        },
        "id": "x9L_O0d7apDw",
        "outputId": "b383665d-bf49-4f9b-eb45-84113b17dd43"
      },
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: Logging into wandb.ai. (Learn how to deploy a W&B server locally: https://wandb.me/wandb-server)\n",
            "\u001b[34m\u001b[1mwandb\u001b[0m: You can find your API key in your browser here: https://wandb.ai/authorize\n",
            "\u001b[34m\u001b[1mwandb\u001b[0m: Paste an API key from your profile and hit enter, or press ctrl+c to quit: \n",
            "\u001b[34m\u001b[1mwandb\u001b[0m: Appending key for api.wandb.ai to your netrc file: /root/.netrc\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\u001b[34m\u001b[1mwandb\u001b[0m: Currently logged in as: \u001b[33mmattslight\u001b[0m (\u001b[33mkpsworld\u001b[0m). Use \u001b[1m`wandb login --relogin`\u001b[0m to force relogin\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Tracking run with wandb version 0.12.18"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Run data is saved locally in <code>/content/wandb/run-20220614_113912-1nf4wrb4</code>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ],
            "text/html": [
              "Syncing run <strong><a href=\"https://wandb.ai/kpsworld/uncategorized/runs/1nf4wrb4\" target=\"_blank\">eternal-brook-17</a></strong> to <a href=\"https://wandb.ai/kpsworld/uncategorized\" target=\"_blank\">Weights & Biases</a> (<a href=\"https://wandb.me/run\" target=\"_blank\">docs</a>)<br/>"
            ]
          },
          "metadata": {}
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<button onClick=\"this.nextSibling.style.display='block';this.style.display='none';\">Display W&B run</button><iframe src=\"https://wandb.ai/kpsworld/uncategorized/runs/1nf4wrb4?jupyter=true\" style=\"border:none;width:100%;height:420px;display:none;\"></iframe>"
            ],
            "text/plain": [
              "<wandb.sdk.wandb_run.Run at 0x7f6673145050>"
            ]
          },
          "metadata": {},
          "execution_count": 1
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 2. Download the MNIST dataset"
      ],
      "metadata": {
        "id": "h0v8WwagAuHs"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7oTkBmSCXZij",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "4e8be988-a89f-4af1-b773-0fd376ea9e08"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/tf-keras-datasets/mnist.npz\n",
            "11493376/11490434 [==============================] - 0s 0us/step\n",
            "11501568/11490434 [==============================] - 0s 0us/step\n"
          ]
        }
      ],
      "source": [
        "from tensorflow.keras.datasets import mnist\n",
        "(train_images, train_labels), (test_images, test_labels) = mnist.load_data()"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 3. Check the size and shape of the dataset"
      ],
      "metadata": {
        "id": "2KSipSkcBDYY"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_images.shape)\n",
        "print(train_labels.shape)\n",
        "\n",
        "print(test_images.shape)\n",
        "print(test_labels.shape)\n",
        "\n",
        "## (60000 = number of training samples, 28 = width of image, 28 = height of image)\n",
        "## (60000 = number of labels, )"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dfbd4oZ2XIIV",
        "outputId": "95161eb4-121a-4b0e-e327-812592269c38"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "(60000, 28, 28)\n",
            "(60000,)\n",
            "(10000, 28, 28)\n",
            "(10000,)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(train_images[0])"
      ],
      "metadata": {
        "id": "_oP5qY6wrKLS"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 4. Build the model"
      ],
      "metadata": {
        "id": "jZfda6LkBHsa"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "from tensorflow import keras \n",
        "from tensorflow.keras import layers\n",
        "\n",
        "\n",
        "'''model = keras.Sequential([\n",
        "    layers.Dense(512, activation=\"relu\"),\n",
        "    layers.Dense(10, activation=\"softmax\")\n",
        "])\n",
        "\n",
        "'''\n",
        "model = keras.Sequential()\n",
        "\n",
        "model.add(layers.Conv2D(filters = 32, kernel_size = (5,5),padding = 'Same', activation ='relu', input_shape = (28,28,1)))\n",
        "model.add(layers.Conv2D(filters = 32, kernel_size = (5,5),padding = 'Same', activation ='relu'))\n",
        "model.add(layers.MaxPool2D(pool_size=(2,2)))\n",
        "model.add(layers.Dropout(0.25))\n",
        "model.add(layers.Conv2D(filters = 64, kernel_size = (3,3),padding = 'Same', activation ='relu'))\n",
        "model.add(layers.Conv2D(filters = 64, kernel_size = (3,3),padding = 'Same', activation ='relu'))\n",
        "model.add(layers.MaxPool2D(pool_size=(2,2), strides=(2,2)))\n",
        "model.add(layers.Dropout(0.25))\n",
        "model.add(layers.Flatten())\n",
        "model.add(layers.Dense(256, activation = \"relu\"))\n",
        "model.add(layers.Dense(10, activation = \"softmax\"))"
      ],
      "metadata": {
        "id": "CaCtiGGXXj8t"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 5. Compile the model"
      ],
      "metadata": {
        "id": "SuJiNf6eBMK2"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.compile(optimizer=\"rmsprop\",\n",
        "              loss=\"sparse_categorical_crossentropy\",\n",
        "              metrics=[\"accuracy\"])"
      ],
      "metadata": {
        "id": "cNRfPqFcXt6t"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "#train_images = train_images.reshape((60000, 28 * 28)) ### Flatten a 28 * 28 image into a 1-d vector\n",
        "train_images = train_images.astype(\"float32\") / 255 \n",
        "#test_images = test_images.reshape((10000, 28 * 28))  ### Flatten a 28 * 28 image into a 1-d vector\n",
        "test_images = test_images.astype(\"float32\") / 255"
      ],
      "metadata": {
        "id": "V1Z4-21HX0Ku"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 6. Train (fit) the model \n"
      ],
      "metadata": {
        "id": "RoVBRTj7Bhhi"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "model.fit(train_images, train_labels, epochs=5, batch_size=128, callbacks=[WandbCallback(log_weights=True)])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "WmO8_jWTX7Ad",
        "outputId": "5be4659b-03b0-4ccc-a63f-5187e879a454"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/5\n",
            "469/469 [==============================] - 17s 13ms/step - loss: 0.1677 - accuracy: 0.9482 - _timestamp: 1655204520.0000 - _runtime: 215.0000\n",
            "Epoch 2/5\n",
            "469/469 [==============================] - 5s 10ms/step - loss: 0.0482 - accuracy: 0.9849 - _timestamp: 1655204525.0000 - _runtime: 220.0000\n",
            "Epoch 3/5\n",
            "469/469 [==============================] - 5s 10ms/step - loss: 0.0345 - accuracy: 0.9894 - _timestamp: 1655204530.0000 - _runtime: 225.0000\n",
            "Epoch 4/5\n",
            "469/469 [==============================] - 5s 10ms/step - loss: 0.0276 - accuracy: 0.9912 - _timestamp: 1655204534.0000 - _runtime: 229.0000\n",
            "Epoch 5/5\n",
            "469/469 [==============================] - 5s 10ms/step - loss: 0.0242 - accuracy: 0.9931 - _timestamp: 1655204539.0000 - _runtime: 234.0000\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<keras.callbacks.History at 0x7fa65626bd50>"
            ]
          },
          "metadata": {},
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "## 7. Test the the model"
      ],
      "metadata": {
        "id": "N7AMRczMBnwv"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "predictions = model.predict(test_images[0:20])\n",
        "predictions = np.argmax(predictions, axis=1)\n",
        "out = np.around(predictions, decimals=2)\n",
        "print(out[0:20])\n",
        "print(test_labels[0:20])"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ZcEUu3PzbXiQ",
        "outputId": "ab49d3c3-8164-48cd-921d-a5f40238e9c5"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[7 2 1 0 4 1 4 9 5 9 0 6 9 0 1 5 9 7 3 4]\n",
            "[7 2 1 0 4 1 4 9 5 9 0 6 9 0 1 5 9 7 3 4]\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_loss, test_acc = model.evaluate(test_images, test_labels)\n",
        "print(f\"test_acc: {test_acc}\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p0TXIlkGcKfr",
        "outputId": "64d2014d-369c-4e44-c4c4-ca6d7ac84e91"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "313/313 [==============================] - 11s 36ms/step - loss: 0.0316 - accuracy: 0.9889\n",
            "test_acc: 0.9889000058174133\n"
          ]
        }
      ]
    }
  ]
}