Classification_CNNs.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "Classification_CNNs.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.6.8"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/raqueeb/TensorFlow2/blob/master/Classification_CNNs.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "jYysdyb-CaWM"
      },
      "source": [
        "# কনভল্যুশনাল নিউরাল নেটওয়ার্ক দিয়ে ইমেজ ক্লাসিফিকেশন "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RsBSPVjuwF6A",
        "colab_type": "text"
      },
      "source": [
        "### শুরুতেই বেসিক ধারণা "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "li1dtIMAwF6B",
        "colab_type": "text"
      },
      "source": [
        "আজকে আমরা আলাপ করছি ‘কনভল্যুশনাল নিউরাল নেটওয়ার্ক’ দিয়ে। একে আমরা সচরাচর ‘সিএনএন’ বলে থাকি। আমাদের এই নতুন ডিপ নিউরাল নেটওয়ার্কের কাজ আগের নেটওয়ার্ক থেকে অনেকটাই দক্ষ, বিশেষ করে ইমেজ প্রসেসিং এ। ব্যাপারটা বোঝার জন্য আমাদেরকে তাকাতে হবে মানুষের মস্তিষ্কের দিকে। আমরা সারাদিন কি করি? সজ্ঞানে অথবা ভুলোমনে হোক, আমরা সারাদিন ধরে নতুন জিনিসপত্র দেখি, সেটা কে চিনে মনে মনে লেবেল করি, এইটা হচ্ছে পাহাড়, এইটা বাঘ - শেষে সবকিছুর একটা প্যাটার্ন বুঝতে পারি। আবার, যেটা জানিনা বা বুঝিনা সেটাই ভূত। \n",
        "\n",
        "ব্যাপারটা কিন্তু এমনিতেই হয়নি। মানুষের প্রায় ৫০ কোটি বছর লেগেছে এই ধরনের জিনিস মাথায় তৈরি হতে। আমাদের চোখ এবং ব্রেনের মধ্যে যে কোঅর্ডিনেশন এবং কোলাবরেশন তার জন্যই কিন্তু আমরা দেখতে পারি পৃথিবীর আশেপাশে। একটা বাচ্চা যেভাবে তার আশেপাশের বিভিন্ন অবজেক্টকে দেখে চিনতে শিখে, সেভাবে আমাদের অ্যালগরিদমকে লক্ষ লক্ষ ছবি দেখাতে হবে যাতে তার একটা 'জেনারেলাইজড' আইডিয়া হয়, কোনটা কি হতে পারে। এরপর একটা ছবি যেটা আগে তাকে দেখানো হয়নি, সেটা দেখে বলতে পারবে ছবিটা কিসের। কম্পিউটার আর মানুষের মধ্যে দেখার পার্থক্য আছে। কম্পিউটার যা দেখে তার সবই সংখ্যা। প্রতিটা ছবি আসলে একটা দুই ডাইমেনশনের সংখ্যার অ্যারে, যেগুলোকে পিক্সেল বলছি আমরা। \n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/start-page/neuron.jpg\"> চিত্রঃ মানুষ কিভাবে দেখে? আমাদের কনভল্যুশনাল নিউরাল নেটওয়ার্কের মতো \n",
        "\n",
        "আজকে যে ‘কনভল্যুশনাল নিউরাল নেটওয়ার্ক’ নিয়ে আলাপ করছি সেটা নিয়ে রিসার্চ হয়েছে ১৯৬০ সালের দিকে। বিশেষ করে মানুষ এবং অন্যান্য স্তন্যপায়ী প্রানীরা কিভাবে দেখে। আমাদের ভিজুয়াল কর্টেক্সে নিউরনগুলো কিভাবে কাজ করে তার থেকেই ধারণা নিয়ে তৈরি এই কনভলিউশনাল নেটওয়ার্ক। সাধারণ নিউরাল নেটওয়ার্ক থেকে এ কারণে এই নেটওয়ার্কের আর্কিটেকচার ভিন্ন। একে অনেকে ডিপ নিউরাল নেটওয়ার্ক বলতে নারাজ। সাধারণ নিউরাল নেটওয়ার্ক এবং কনভল্যুশনাল নিউরাল নেটওয়ার্কের প্রাথমিক পার্থক্য হচ্ছে এদের কাজের ধারা, এবং একটা কানেক্টেড, আরেকটা অতোটা কানেক্টেড না। কনভলিউশন শব্দটা এসেছে অংক থেকে যার মানে হচ্ছে দুটো ফাংশনের একটা ম্যাথমেটিক্যাল কম্বিনেশন যেটার আউটকাম তৃতীয় একটা ফাংশন। ইনপুট থেকে পাওয়া দুটো সেটকে এক জায়গায় নিয়ে আসাকে কনভলিউশন বলে।\n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/start-page/cnn2.png\"> চিত্রঃ কনভল্যুশনাল নিউরাল নেটওয়ার্কের একটা এন্ড টু এন্ড ডায়াগ্রাম "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YyE4uqh6wF6C",
        "colab_type": "text"
      },
      "source": [
        "### ফিচার এক্সট্রাকশন এবং ক্লাসিফিকেশন "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "sIdmW8UVwF6D",
        "colab_type": "text"
      },
      "source": [
        "কনভল্যুশনাল নিউরাল নেটওয়ার্কের দুটো ভাগ\n",
        "\n",
        "১. ফিচার এক্সট্রাকশন/হিডেন লেয়ার পার্ট: এখানে কনভল্যূশন এবং পুলিং (পরে বলছি) এর মাধ্যমে ফিচার এক্সট্রাকশন মানে মানুষের মুখ হলে তার নাক চোখ, মুখ এগুলোকে ডিটেক্ট করে আগে লেয়ার বাই  লেয়ার। এর কাজ শেষ হলে পরেরটা শুরু হয়। \n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cnn7.jpeg\"> চিত্রঃ কনভল্যুশনাল নিউরাল নেটওয়ার্কের শুরুর অংশ \n",
        "\n",
        "২. এটা ফুলি কানেক্টেড লেয়ার, যা কাজ করে ফিচার এক্সট্রাক্ট করার হয়ে গেলে - ক্লাসিফাইয়ার তার প্রোবাবিলিটি ডিস্ট্রিবিউশন থেকে যার ভোটিং বেশি হয় সেই ছবিকে বলে দেয়। \n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/neural_net2.jpeg\"> চিত্রঃ সাধারণ নিউরাল নেটওয়ার্কের শেষের অংশ \n",
        "\n",
        "সাধারণ নিউরাল নেটওয়ার্কের ইনপুট এ যা দেয়া হয় সেটাকে সে ট্রান্সফরম করতে থাকে তারপরের হিডেন লেয়ারের সিরিজ থেকে। প্রতিটা লেয়ারে অনেকগুলো নিউরন থাকে, আবার প্রতিটা লেয়ার একটা আরেকটার সাথে 'ফুলি কানেক্টেড'। সেখানে কনভল্যুশনাল নিউরাল নেটওয়ার্কগুলো একটু ভিন্ন। তাদের লেয়ারগুলোকে আমরা অরগানাইজ করি তিন ডাইমেনশনে মানে দৈর্ঘ্য প্রস্থ এবং উচ্চতায় (ছবি হলে তিন রঙয়ের চ্যানেল)। ভয় পাবেন না, সামনে আমরা সেগুলো ছবি দিয়ে বুঝিয়ে দেব। এখানে সব নিউরনগুলো একটা আরেকটার সাথে কানেক্টেড না, তবে ছোট ছোট রিজিয়ন দিয়ে কানেক্টেড। এটাই সিএনএনের বড় প্লাস পয়েন্ট। \n",
        "\n",
        "সাধারণ নিউরাল নেটওয়ার্ক এ আমরা পুরো ছবিকেই প্রসেস করতাম। এখানে একটা ছবি অনেকগুলো ফিল্টারের মধ্যে দিয়ে যাবে। শুরুতে আমাদের এখানে ইনপুট ডাটা ছোট একটা ফিল্টার এর মধ্যে যাবে, এখানে ফিল্টার অথবা কার্নাল এই দুটো আসল ইনপুট ইমেজ থেকে ছোট তাই এর দক্ষতা অনেক ভালো। এই ফিল্টারগুলোর আউটপুট জমা হয় ফিচার ম্যাপে। বড় একটা ছবিকে তার ছোট পিক্সেল সাইজ ফিল্টার দিয়ে যেই ম্যাথমেটিক্যাল অপারেশন করবো তার আউটপুট এসে জমা হবে ফিচার ম্যাপে। এই ফিল্টারগুলো স্লাইডিং করবে পুরো ইমেজ জুড়ে। ইমেজ এবং ফিল্টারের মাট্রিক্স মালটিপ্লিকেশন হবার পর সেটার যোগফল এসে জমা হয় ফিচার ম্যাপে। আবারো বলছি কনভলিউশন হচ্ছে অরিজিনাল ছবির ডাটা সাথে ফিল্টারের মাট্রিক্স মালটিপ্লিকেশন এবং তার রেজাল্ট এসে ধুঁকছে ফিচার ম্যাপে। এর পাশাপাশি এই আউটপুটকে আমরা বলছি ডাউন স্যাম্পলিং, যেটাকে সবাই বলছে 'ম্যাক্স পুলিং' যার কাজ হচ্ছে আমরা যখন একটা রিজিওন সিলেক্ট করি তখন সেখান থেকে সর্বোচ্চ ভ্যালুটাকে আমরা সেই রিজিওনের জন্য নতুন ভ্যালু হিসেবে ডিক্লেয়ার করি। আসুন ছবি দেখে হিসেব মিলাই। এখানে একটা ছবি থেকে বিড়ালকে আমরা ক্লাসিফাই করব। এই মজার ছবি আর কনটেন্ট এর জন্য হ্যারিসনকে ধন্যবাদ। উনার সাইট হচ্ছে pythonprogramming.net ।"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y-j3T3cVwF6E",
        "colab_type": "text"
      },
      "source": [
        "## ছবিতে বিড়ালের ফিচার এক্সট্রাকশন এবং ক্লাসিফিকেশন\n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat1.png\"> চিত্রঃ একটা বেড়াল, বেড়াল মনে হচ্ছে না? \n",
        "\n",
        "১. এক নাম্বার ছবিতে আমরা ধরে নিচ্ছি একটা বিড়াল আছে। বিড়ালটার আঁকাটা যতই খারাপ হোক না কেন আমার মনে হয় বোঝা যাচ্ছে এটা যে একটা বিড়াল। \n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat2.png\"> চিত্রঃ ৫ X ৪ পিক্সেল ছবি \n",
        "\n",
        "২. মেশিনের বোঝার সুবিধার জন্য এই চারকোনা ঘরগুলো মানে একেকটা পিক্সেল দিয়ে ভাগ করে ফেলেছি। একটা ছবিতে মিলিয়ন পিক্সেল থাকতে পারে তবে আমরা বোঝার সুবিধার জন্য এখানে কমিয়ে দিয়েছি ৫ X ৪ পিক্সেলে।\n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat3.png\"> চিত্র:৫ X ৪ পিক্সেলের ছবি \n",
        "\n",
        "৩. কনভল্যুশন এর সময় আমরা এই বিড়ালের ছবিটার পুরো অংশ না নিয়ে একটা অংশ নেব। এখানে আপনারা দেখছেন একটা ৩ X ৩ একটা উইন্ডো নিয়েছি এই ফিল্টার বানানোর জন্য। আমরা এই ৩ X ৩ পিক্সেলের রিজিওন এর মধ্যেই আমাদের ফিচার খুজবো। এই ৩ X ৩ পিক্সেলের আউটপুট আমরা নিয়ে আসব ১ (এক) পিক্সেলের ফিচার ম্যাপে। এখানে আমরা একটা পিক্সেলের ফিচার ম্যাপ দেখাচ্ছি, আসল সময় এখানে অনেকগুলো ফিচার ম্যাপ আসবে।\n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat4.png\"> চিত্রঃ একটা পিক্সেলের ফিচার ম্যাপ\n",
        "\n",
        "৪. এখন আমাদের ৩ X ৩ যে কার্নাল বা ফিল্টার ব্যবহার করেছি, সেটাকে ডানে ১ পিক্সেল করে স্লাইড করব। এরপর আরো ১ পিক্সেল করে স্লাইড করবো। এর অর্থ ২ পিক্সেলের স্ট্রাইড। মনে রাখতে হবে যাতে আমরা কোন পিক্সেল এই ফিল্টার থেকে হাতছাড়া না হয়ে পড়ে। এইজন্য অনেক পিক্সেল ওভারল্যাপ হবে, কিন্তু এতে কোন পিক্সেল বাদ পড়বে না। \n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat5.png\"> চিত্রঃ ২ পিক্সেলের স্ট্রাইড\n",
        "\n",
        "৫. এইভাবে আমরা ডানে বামে উপরে নিচে সবজায়গায় স্লাইডিং করব, যাতে পুরো ইমেজটাকে কাভার করতে পারি। আমাদের এই ৩ X ৩ ফিল্টার দিয়ে যে ফিচার ম্যাপ তৈরি করছি সেটা ৩ X ৩ ফিল্টার থেকে একটা একটা করে ইনপুট নিয়ে ফিচার ম্যাপ তৈরি করেছে। অনেকগুলো পিকচার থেকে আমরা এখন কমিয়ে নিয়ে এলাম মাত্র চারটা পিক্সেল ভ্যালুর ফিচার ম্যাপ।\n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat6.png\"> চিত্রঃ চারটা পিক্সেল ভ্যালুর ফিচার ম্যাপ\n",
        "\n",
        "৬. এখন আমরা পুলিং এর কাজ করব। আমরা ধরে নিচ্ছি আমাদের কনভলিউশন ছবির বদলে এই সংখ্যাগুলো দিয়েছে। এই সংখ্যাগুলো এসেছে পিক্সএল ভ্যালু এবং ফিল্টার ভ্যালু এর মাট্রিক্স মালটিপ্লিকেশন থেকে।\n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat7.png\"> চিত্রঃ পুলিং এর কাজ\n",
        "\n",
        "৭. এখানে আমরা আবার একটা ৩ X ৩ পুলিং উইন্ডো নিচ্ছি। \n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat8.png\"> চিত্রঃ ৩ X ৩ পুলিং উইন্ডো\n",
        "\n",
        "৮. এই পুলিং এর সবচেয়ে বেশি ব্যবহৃত ফর্ম হচ্ছে 'ম্যাক্স পুলিং', যার কাজ হচ্ছে ওই উইন্ডোর মধ্যে থেকে সবচেয়ে বেশি ভ্যালুকে বের করে আনা। ওই উইন্ডো বা রিজিওনের জন্য সেটাই নতুন ভ্যালু। আপনি দেখুন প্রথম উইন্ডোতে ৩ X ৩ পুলিং ৪ হচ্ছে সর্বোচ্চ সংখ্যা। \n",
        "\n",
        "<img src=\"https://raw.githubusercontent.com/raqueeb/deep_learning_book/master/assets/cat9.png\"> চিত্রঃ সবগুলো পিক্সেল থেকে নতুন চারটা ভ্যালু\n",
        "\n",
        "৯. দুই পিক্সেল ডানে নতুন যে উইন্ডো হল তার সর্বোচ্চ ভ্যালু হচ্ছে ৩। উপর থেকে নিচের দিকে ১ পিক্সেল স্লাইড করলে যে ৩ X ৩ উইন্ডো পাওয়া যায় তার সর্বোচ্চ ভ্যালু হচ্ছে ৪। হিসেবে নিচের উইন্ডোকে আরো দুই পিক্সেল ডানে নিলে যে ম্যাক্স পুলিং উইন্ডো হবে সেটার সর্বোচ্চ ভ্যালু হচ্ছে ৩। আমরা সবদিকে স্লাইডিং করে পুরনো ছবি তে সবগুলো পিক্সেল কাভার করলে আমরা নতুন চারটা ভ্যালু পাব যা রিপ্রেজেন্ট করবে বিড়ালের পুরো ছবিটাকে।"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "mqd_yq67wF6F",
        "colab_type": "text"
      },
      "source": [
        "বেশি ঝামেলার মনে হচ্ছে? আচ্ছা, আপাততঃ কনভল্যুশনাল নিউরাল নেটওয়ার্ক দিয়ে ইমেজ ক্লাসিফিকেশন না করে কিভাবে সেই ডেটাসেট সহজে ব্যবহার করা যায় সেটার একটা পন্থা বের করছি। এরপর ফিরে যাবো কনভল্যুশনাল নিউরাল নেটওয়ার্ক দিয়ে কাজে। "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "H0tMfX2vR0uD"
      },
      "source": [
        "## ইনস্টল এবং ইমপোর্ট করে নেই দরকারি লাইব্রেরি "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NA2ZzURQwF6G",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import math\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "5HDhfftMGc_i",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "c465b392-36b8-4b65-ba82-87e3883649ac"
      },
      "source": [
        "try:\n",
        "  # শুধুমাত্র টেন্সর-ফ্লো ২.x ব্যবহার করবো \n",
        "  %tensorflow_version 2.x\n",
        "except Exception:\n",
        "  pass\n",
        "\n",
        "import tensorflow as tf"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "TensorFlow 2.x selected.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qtvAGEwUwF6N",
        "colab_type": "text"
      },
      "source": [
        "## টেন্সর-ফ্লো ডেটাসেট এপিআই ব্যবহার\n",
        "\n",
        "মনে আছে সাইকিট লার্নের কথা? এর ভেতরের ডেটাসেট নিয়ে কাজ করেছিলাম আমরা। আলাদা করে ডাউনলোড করতে হয়নি। আমরা সেভাবেই টেন্সর-ফ্লো ডেটাসেট এপিআই [TensorFlow Datasets](https://www.tensorflow.org/datasets/), ব্যবহার করবো যাতে এই বিশ্বখ্যাত সব ডেটাসেটে আমরা এক্সেস পাই সহজে, ঝামেলাপূর্ণ ডাউনলোড ছাড়াই - শুধু জানতে হবে সেটার কনভেনশন। "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AGzIIZolwF6P",
        "colab_type": "text"
      },
      "source": [
        "### আমাদের সব ডেটাসেট এক্সপোজ করা আছে tf.data.Datasets এ \n",
        "কিছু টেস্ট চালাই এখানে "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_md9NvmrwF6Q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# টেন্সর-ফ্লো ডেটাসেট tfds\n",
        "import tensorflow_datasets as tfds"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cIwTb6xcwF6S",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 55
        },
        "outputId": "959dc7b6-384b-4f5a-d69b-53da143236ee"
      },
      "source": [
        "# আপনি দেখুন কতো কতো রিসার্চ ডেটাসেট এখানে \n",
        "# প্রতিটা ডেটাসেট ইমপ্লিমেন্ট করা আছে  tfds.core.DatasetBuilder হিসেবে \n",
        "# তার আগে লিস্ট দেখি \n",
        "print(tfds.list_builders())"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['abstract_reasoning', 'aflw2k3d', 'amazon_us_reviews', 'bair_robot_pushing_small', 'bigearthnet', 'binarized_mnist', 'binary_alpha_digits', 'caltech101', 'caltech_birds2010', 'caltech_birds2011', 'cars196', 'cats_vs_dogs', 'celeb_a', 'celeb_a_hq', 'chexpert', 'cifar10', 'cifar100', 'cifar10_corrupted', 'clevr', 'cmaterdb', 'cnn_dailymail', 'coco', 'coco2014', 'coil100', 'colorectal_histology', 'colorectal_histology_large', 'curated_breast_imaging_ddsm', 'cycle_gan', 'deep_weeds', 'definite_pronoun_resolution', 'diabetic_retinopathy_detection', 'downsampled_imagenet', 'dsprites', 'dtd', 'dummy_dataset_shared_generator', 'dummy_mnist', 'emnist', 'eurosat', 'fashion_mnist', 'flores', 'food101', 'gap', 'glue', 'groove', 'higgs', 'horses_or_humans', 'image_label_folder', 'imagenet2012', 'imagenet2012_corrupted', 'imdb_reviews', 'iris', 'kitti', 'kmnist', 'lfw', 'lm1b', 'lsun', 'malaria', 'mnist', 'mnist_corrupted', 'moving_mnist', 'multi_nli', 'multi_nli_mismatch', 'nsynth', 'omniglot', 'open_images_v4', 'oxford_flowers102', 'oxford_iiit_pet', 'para_crawl', 'patch_camelyon', 'pet_finder', 'places365_small', 'quickdraw_bitmap', 'resisc45', 'rock_paper_scissors', 'rock_you', 'scene_parse150', 'shapes3d', 'smallnorb', 'snli', 'so2sat', 'squad', 'stanford_dogs', 'stanford_online_products', 'starcraft_video', 'sun397', 'super_glue', 'svhn_cropped', 'ted_hrlr_translate', 'ted_multi_translate', 'tf_flowers', 'the300w_lp', 'titanic', 'trivia_qa', 'uc_merced', 'ucf101', 'visual_domain_decathlon', 'voc2007', 'wikipedia', 'wmt14_translate', 'wmt15_translate', 'wmt16_translate', 'wmt17_translate', 'wmt18_translate', 'wmt19_translate', 'wmt_t2t_translate', 'wmt_translate', 'xnli']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TzuLj-bzwF6V",
        "colab_type": "text"
      },
      "source": [
        "## আমরা একটু দেখি tfds.load এবং DatasetBuilder এর কাজ "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "B33RzDB0wF6W",
        "colab_type": "text"
      },
      "source": [
        "### `tfds.load`: যাকে আমরা বলছি এক লাইনে এক ডেটাসেট \n",
        "\n",
        "`tfds.load` দিয়ে খুব সহজে একটা ডেটাসেট বিল্ড এবং লোড করা যায়। `tf.data.Dataset` হচ্ছে স্ট্যান্ডার্ড টেন্সর-ফ্লো এপিআই বিশেষ করে ইনপুট পাইপলাইন তৈরি করতে। আমরা ডাউনলোড করে নিতে পারি `data_dir=`তে বলে। আর তা না হলে ডিফল্ট হচ্ছে `~/tensorflow_datasets/`)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Qcaz5_ebwF6X",
        "colab_type": "code",
        "colab": {
          "resources": {
            "http://localhost:8080/nbextensions/google.colab/colabwidgets/controls.css": {
              "data": "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",
              "ok": true,
              "headers": [
                [
                  "content-type",
                  "text/css"
                ]
              ],
              "status": 200,
              "status_text": ""
            }
          },
          "base_uri": "https://localhost:8080/",
          "height": 532
        },
        "outputId": "54476cb7-72b7-4a47-f93a-42969180e705"
      },
      "source": [
        "# এই mnist কিন্তু হাতে লেখা সংখ্যাকে চেনার ডেটাসেট, ফ্যাশন নয় \n",
        "# একঘেঁয়েমি কাটানোর জন্য নতুন জিনিস \n",
        "mnist_train = tfds.load(name=\"mnist\", split=\"train\")\n",
        "assert isinstance(mnist_train, tf.data.Dataset)\n",
        "print(mnist_train)"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[1mDownloading and preparing dataset mnist (11.06 MiB) to /root/tensorflow_datasets/mnist/1.0.0...\u001b[0m\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "9c30fe29fd2c4ce9bcefd645473ef633",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Dl Completed...', max=1, style=ProgressStyl…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "632e5dbe3616446a92458efdb27d7f50",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Dl Size...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "9dc60b0181e64a50ab5ba5195775e6c7",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Extraction completed...', max=1, style=Prog…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/urllib3/connectionpool.py:847: InsecureRequestWarning: Unverified HTTPS request is being made. Adding certificate verification is strongly advised. See: https://urllib3.readthedocs.io/en/latest/advanced-usage.html#ssl-warnings\n",
            "  InsecureRequestWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/urllib3/connectionpool.py:847: InsecureRequestWarning: Unverified HTTPS request is being made. Adding certificate verification is strongly advised. See: https://urllib3.readthedocs.io/en/latest/advanced-usage.html#ssl-warnings\n",
            "  InsecureRequestWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/urllib3/connectionpool.py:847: InsecureRequestWarning: Unverified HTTPS request is being made. Adding certificate verification is strongly advised. See: https://urllib3.readthedocs.io/en/latest/advanced-usage.html#ssl-warnings\n",
            "  InsecureRequestWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/urllib3/connectionpool.py:847: InsecureRequestWarning: Unverified HTTPS request is being made. Adding certificate verification is strongly advised. See: https://urllib3.readthedocs.io/en/latest/advanced-usage.html#ssl-warnings\n",
            "  InsecureRequestWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "\n",
            "\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "9bf9c3cd53b24dafb5b2d15655389228",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\r"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "eb6e49ce9c7a4649843b030075ab9ce2",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Shuffling...', max=10, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_datasets/core/file_format_adapter.py:209: tf_record_iterator (from tensorflow.python.lib.io.tf_record) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use eager execution and: \n",
            "`tf.data.TFRecordDataset(path)`\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow_datasets/core/file_format_adapter.py:209: tf_record_iterator (from tensorflow.python.lib.io.tf_record) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use eager execution and: \n",
            "`tf.data.TFRecordDataset(path)`\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "c973942a796a4801a04e054369c5f8a6",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "7bee185fc5c246db93fa2d342fb5149a",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "ac62026da0934289a7ffcac23122791a",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "cbbccb8cb65d43d5bd1c6f52a6c45ef0",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "1d0ef5b255a242fba138a69496f352d4",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "7845d653789043b3927d798eb52f1c64",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "ef9fb09e1fbf476cbcb2eb52df73f127",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "392822129f1f41b18a036d04f9029fe9",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "9cc97194baa44585ae52a4348ffa4e31",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "422815ed70434a55bb730c73af61c6cd",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "364a3e1e8c024b428d16dad2b5d7c651",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "2f689a1b373a4395a2d6ad3c8a93220e",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "b5cce748fa1949bab06ea81798d13d69",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "f88a7c3e53c749c2a492b18d627f7070",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "629e2ccd9eda41bf92c374f5ee46ceea",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "fc38008d981f4179813c89304d9a1a17",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "a5660951088949a78f03a64976f131a3",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "ef3ec5b39f38420388b7254aaac0f1ff",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "38d0b568ad244eb1953ecb130d36c05e",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "0aeffd110774470fa77d56263709c76e",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=6000, style=ProgressStyle(description_width=…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\r"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "024bc9cba58c4afc91c0bd7b8fbf0b32",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', max=1), HTML(value='')))"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\r"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "b1178cabfc4b4004b825d63b6705d108",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Shuffling...', max=1, style=ProgressStyle(description_width='…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "f37ebdc3925c42e4a189fe9612572385",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=1, bar_style='info', description='Reading...', max=1, style=ProgressStyle(des…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "b40daddd309c4413a95435eacdbcc028",
              "version_minor": 0,
              "version_major": 2
            },
            "text/plain": [
              "HBox(children=(IntProgress(value=0, description='Writing...', max=10000, style=ProgressStyle(description_width…"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "\r\u001b[1mDataset mnist downloaded and prepared to /root/tensorflow_datasets/mnist/1.0.0. Subsequent calls will reuse this data.\u001b[0m\n",
            "<_OptionsDataset shapes: {image: (28, 28, 1), label: ()}, types: {image: tf.uint8, label: tf.int64}>\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5_3I9kbQwF6a",
        "colab_type": "text"
      },
      "source": [
        "### কোন ভার্সনের ডেটাসেট দরকার?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RL1UlFrCwF6b",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "mnist = tfds.load(\"mnist:1.*.*\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "DEG8p0WGwF6e",
        "colab_type": "text"
      },
      "source": [
        "### ফিচার ডিকশনারি \n",
        "\n",
        "আমাদের সাইকিট লার্নের মতো এখানে টেন্সর-ফ্লো ডেটাসেট tfds এর ফিচার ডিকশনারির সব ফিচারকে ম্যাপ করা হয়েছে টেন্সর ভ্যালুতে। যেমন, আমাদের ডেটাসেট MNIST, এর দুটো কী আছে: \"image\" এবং \"label\"। আমরা একটা রেকর্ড মানে ফুল এক্সাম্প্ল দেখি। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SrCzbzyfwF6f",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 287
        },
        "outputId": "40f4e142-85b3-45cd-91ac-f22b11db8439"
      },
      "source": [
        "for mnist_example in mnist_train.take(1):  # একটা উদাহরণ নেই (একটা জোড়া)\n",
        "  image, label = mnist_example[\"image\"], mnist_example[\"label\"]\n",
        "\n",
        "  plt.imshow(image.numpy()[:, :, 0].astype(np.float32), cmap=plt.get_cmap(\"gray\"))\n",
        "  print(\"Label: %d\" % label.numpy())"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Label: 3\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAP8AAAD8CAYAAAC4nHJkAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "zSYcEB44wF6i",
        "colab_type": "text"
      },
      "source": [
        "## DatasetBuilder এর কাজ \n",
        "\n",
        "DatasetBuilder আসলে tfds.load কে সাহায্য করে অন্যভাবে, একই ধরণের কাজ কিছুটা। আমরা আগের কাজটা করে দেখি MNIST DatasetBuilder দিয়ে। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "drR0MJpDwF6j",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "mnist_builder = tfds.builder(\"mnist\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "pe5rRFlowF6m",
        "colab_type": "text"
      },
      "source": [
        "### ইনফো এট্রিবিউট দেখি, সব তথ্য আছে এখানে "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qYwcaldxwF6n",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 451
        },
        "outputId": "ca3ab1af-927c-4043-f0f3-e9fdb44640cb"
      },
      "source": [
        "print(mnist_builder.info)"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tfds.core.DatasetInfo(\n",
            "    name='mnist',\n",
            "    version=1.0.0,\n",
            "    description='The MNIST database of handwritten digits.',\n",
            "    urls=['https://storage.googleapis.com/cvdf-datasets/mnist/'],\n",
            "    features=FeaturesDict({\n",
            "        'image': Image(shape=(28, 28, 1), dtype=tf.uint8),\n",
            "        'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),\n",
            "    }),\n",
            "    total_num_examples=70000,\n",
            "    splits={\n",
            "        'test': 10000,\n",
            "        'train': 60000,\n",
            "    },\n",
            "    supervised_keys=('image', 'label'),\n",
            "    citation=\"\"\"@article{lecun2010mnist,\n",
            "      title={MNIST handwritten digit database},\n",
            "      author={LeCun, Yann and Cortes, Corinna and Burges, CJ},\n",
            "      journal={ATT Labs [Online]. Available: http://yann. lecun. com/exdb/mnist},\n",
            "      volume={2},\n",
            "      year={2010}\n",
            "    }\"\"\",\n",
            "    redistribution_info=,\n",
            ")\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Gcj4uAhTwF6q",
        "colab_type": "text"
      },
      "source": [
        "### ডেটাসেট.ইনফো এর মধ্যে ফিচার নিয়ে অনেক তথ্য আছে "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PUswaC9ewF6q",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        },
        "outputId": "63a48330-16e2-402d-d378-56114f1c4d78"
      },
      "source": [
        "test = mnist_builder.info\n",
        "print(test.features)\n",
        "print(test.features[\"label\"].num_classes)\n",
        "print(test.features[\"label\"].names)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "FeaturesDict({\n",
            "    'image': Image(shape=(28, 28, 1), dtype=tf.uint8),\n",
            "    'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),\n",
            "})\n",
            "10\n",
            "['0', '1', '2', '3', '4', '5', '6', '7', '8', '9']\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "o-60vVMkwF6t",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# যদি আমরা ডাউনলোড এবং সেটাকে তৈরি করতে চাই \n",
        "# mnist_builder.download_and_prepare()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-LES3H6SwF6w",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# পরীক্ষা করে দেখুন \n",
        "# mnist_train = mnist_builder.as_dataset(split=\"train\")\n",
        "# mnist_train"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "_YTuBNPmwF6y",
        "colab_type": "text"
      },
      "source": [
        "### সরাসরি লোড করে দেখি \n",
        "DatasetInfo লোড করি tfds.load দিয়ে with_info=True সহ। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jbdvT5ZCwF6z",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# mnist_train, info = tfds.load(\"mnist\", split=\"train\", with_info=True)\n",
        "mnist_test, info = tfds.load(\"mnist\", split=\"test\", with_info=True)\n",
        "\n",
        "# imdb, info = tfds.load(\"mnist\", with_info=True, as_supervised=True)\n",
        "# train_data, test_data = imdb['train'], imdb['test']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kQAJss1TwF62",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 538
        },
        "outputId": "f2ced833-31cc-4ffc-e713-ecf03c0abf71"
      },
      "source": [
        "fig = tfds.show_examples(info, mnist_test)"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAggAAAIJCAYAAADAoMXGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nFAgAAED8f3jf9UOBp5KVAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 648x648 with 9 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "g3dYgdLUwF64",
        "colab_type": "text"
      },
      "source": [
        "অনেক গল্প হলো, ফিরে আসি কনভল্যুশনাল নিউরাল নেটওয়ার্ক দিয়ে ইমেজ ক্লাসিফিকেশন নিয়ে। কোনটা করবো MNIST না ফ্যাশন MNIST? আমরা নতুন যুগের ফ্যাশনেবল মানুষ। "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "yR0EdgrLCaWR"
      },
      "source": [
        "##  টেন্সর-ফ্লো ডেটাসেট এপিআই ব্যবহার করে নিয়ে আসি Fashion MNIST\n",
        "\n",
        "মনে রাখি একটা কথা। আমরা এই এক্সারসাইজ করেছি সাধারণ নিউরাল নেটওয়ার্ক দিয়ে। আর সেকারণে এটার কনভল্যুশনাল নিউরাল নেটওয়ার্ক নিয়ে আলাপ করবো এখানে। আগের এক্সারসাইজের কোড চলবে এখানে। তাই, ডুপ্লিকেশন করছি না। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "uusvhUp9Gg37",
        "colab": {}
      },
      "source": [
        "# আর বলতে হবে? টেন্সর-ফ্লো ডেটাসেট এপিআই\n",
        "import tensorflow_datasets as tfds\n",
        "tfds.disable_progress_bar()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Q47uwxnzwF6_",
        "colab_type": "text"
      },
      "source": [
        "### আমাদের দরকার train_dataset এবং test_dataset, ব্যবহার করছি tfds.load"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "7MqDQO0KCaWS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 72
        },
        "outputId": "f550109f-ee13-431b-a2e2-ff8fe5e09291"
      },
      "source": [
        "dataset, metadata = tfds.load('fashion_mnist', as_supervised=True, with_info=True)\n",
        "train_dataset, test_dataset = dataset['train'], dataset['test']"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\u001b[1mDownloading and preparing dataset fashion_mnist (29.45 MiB) to /root/tensorflow_datasets/fashion_mnist/1.0.0...\u001b[0m\n",
            "\u001b[1mDataset fashion_mnist downloaded and prepared to /root/tensorflow_datasets/fashion_mnist/1.0.0. Subsequent calls will reuse this data.\u001b[0m\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "SpYmMmH8wF7E",
        "colab_type": "text"
      },
      "source": [
        "### আমরা একটু মেটাডেটা দেখি, কি বলছে ডেটাসেট সম্পর্কে?"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WGg3nFUmwF7F",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 627
        },
        "outputId": "522408ef-82b0-4e0d-e187-30baeb18f8d7"
      },
      "source": [
        "print(metadata)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "tfds.core.DatasetInfo(\n",
            "    name='fashion_mnist',\n",
            "    version=1.0.0,\n",
            "    description='Fashion-MNIST is a dataset of Zalando's article images consisting of a training set of 60,000 examples and a test set of 10,000 examples. Each example is a 28x28 grayscale image, associated with a label from 10 classes.',\n",
            "    urls=['https://github.com/zalandoresearch/fashion-mnist'],\n",
            "    features=FeaturesDict({\n",
            "        'image': Image(shape=(28, 28, 1), dtype=tf.uint8),\n",
            "        'label': ClassLabel(shape=(), dtype=tf.int64, num_classes=10),\n",
            "    }),\n",
            "    total_num_examples=70000,\n",
            "    splits={\n",
            "        'test': 10000,\n",
            "        'train': 60000,\n",
            "    },\n",
            "    supervised_keys=('image', 'label'),\n",
            "    citation=\"\"\"@article{DBLP:journals/corr/abs-1708-07747,\n",
            "      author    = {Han Xiao and\n",
            "                   Kashif Rasul and\n",
            "                   Roland Vollgraf},\n",
            "      title     = {Fashion-MNIST: a Novel Image Dataset for Benchmarking Machine Learning\n",
            "                   Algorithms},\n",
            "      journal   = {CoRR},\n",
            "      volume    = {abs/1708.07747},\n",
            "      year      = {2017},\n",
            "      url       = {http://arxiv.org/abs/1708.07747},\n",
            "      archivePrefix = {arXiv},\n",
            "      eprint    = {1708.07747},\n",
            "      timestamp = {Mon, 13 Aug 2018 16:47:27 +0200},\n",
            "      biburl    = {https://dblp.org/rec/bib/journals/corr/abs-1708-07747},\n",
            "      bibsource = {dblp computer science bibliography, https://dblp.org}\n",
            "    }\"\"\",\n",
            "    redistribution_info=,\n",
            ")\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "t9FDsUlxCaWW"
      },
      "source": [
        "আমাদের এপিআই ফেরৎ দিচ্ছে train_dataset` এবং test_dataset`, আর আগের মেটাডেটা। \n",
        "\n",
        "* মডেলকে ট্রেনিং করবো `train_dataset` দিয়ে। \n",
        "* মডেলকে টেস্ট করবো `test_dataset` দিয়ে। \n",
        "\n",
        "আগেও বলেছি - আমাদের ছবিগুলো ২৮x২৮ পিক্সেলের নামপাই অ্যারে যার গ্রেস্কেলের কালার ইনটেনসিটি ভ্যালু ০ থেকে ২৫৫ এর মধ্যে। এখানে লেবেলগুলো, মানে যাদেরকে আমরা প্রেডিক্ট করবো তাদের ইন্টেজারের অ্যারে, ০ থেকে ৯ পর্যন্ত।"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "IjnLH5S2CaWx",
        "colab": {}
      },
      "source": [
        "class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
        "               'Sandal',      'Shirt',   'Sneaker',  'Bag',   'Ankle boot']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Brm0b_KACaWX"
      },
      "source": [
        "### আমাদের ডেটাগুলোকে এক্সপ্লোর করি\n",
        "\n",
        "এর আগেও আমরা ডাটাগুলোকে এক্সপ্লোর করেছি, তবে এবার আমরা জানতে চাচ্ছি একটু ভিন্নভাবে। একই জিনিস যে কয়েকভাবে করা যায় সেটার একটা উদাহরণ টানছি এখানে। আগের মতো আমরা ৬০,০০০ ছবি ট্রেনিংসেটে এবং ১০,০০০ ছবি টেস্টসেটে আছে সেটাই দেখানোর চেষ্টা করছি।"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "MaOTZxFzi48X",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "95345000-2b8e-4cc5-afaa-e3917b215005"
      },
      "source": [
        "num_train_examples = metadata.splits['train'].num_examples\n",
        "num_test_examples = metadata.splits['test'].num_examples\n",
        "print(\"Number of training examples: {}\".format(num_train_examples))\n",
        "print(\"Number of test examples:     {}\".format(num_test_examples))"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of training examples: 60000\n",
            "Number of test examples:     10000\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ES6uQoLKCaWr"
      },
      "source": [
        "## আমাদের ডাটাকে নর্মালাইজ করছি\n",
        "\n",
        "যেহেতু ছবির ডাটা, সে কারণে ছবির গ্রেস্কেলে ইনটেনসিটি ০ থেকে ২৫৫ ভ্যালুকে আমরা ০ থেকে ১ এর মধ্যে নিয়ে আসব যাতে মেশিন লার্নিং মডেল ঠিকমতো কাজ করতে পারে। এই নর্মালাইজ ফাংশনকে আমরা টেস্ট এবং ট্রেনিং দুটো ডাটা সেটেই ব্যবহার করব।"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "nAsH3Zm-76pB",
        "colab": {}
      },
      "source": [
        "# এখানে ম্যাপ ফাংশনকে ব্যবহার করছি দুটো ডেটাসেটে\n",
        "# শুরুতেই একটা ফাংশন ডিফাইন করি \n",
        "def normalize(images, labels):\n",
        "  images = tf.cast(images, tf.float32)\n",
        "  images /= 255\n",
        "  return images, labels\n",
        "\n",
        "train_dataset =  train_dataset.map(normalize)\n",
        "test_dataset  =  test_dataset.map(normalize)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "59veuiEZCaW4"
      },
      "source": [
        "## তৈরি করছি কনভল্যুশনাল নিউরাল নেটওয়ার্ক \n",
        "\n",
        "শুরুতেই নেটওয়ার্ক আর লেয়ারের কনফিগারশন, তারপর মডেল কম্পাইলেশন। "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "Gxg1XGm0eOBy"
      },
      "source": [
        "### লেয়ারের সেটআপ\n",
        "\n",
        "যেহেতু একেকটার লেয়ার আরেকটা লেয়ারের সাথে সিকুয়েন্সিয়াল, সে কারণেই শুরুতেই tf.keras.Sequential ব্যবহার করছি। বাকিটা আগের মতো। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "9ODch-OFCaW4",
        "colab": {}
      },
      "source": [
        "model = tf.keras.Sequential([\n",
        "    tf.keras.layers.Conv2D(32, (3,3), padding='same', activation=tf.nn.relu,\n",
        "                           input_shape=(28, 28, 1)),\n",
        "    tf.keras.layers.MaxPooling2D((2, 2), strides=2),\n",
        "    tf.keras.layers.Conv2D(64, (3,3), padding='same', activation=tf.nn.relu),\n",
        "    tf.keras.layers.MaxPooling2D((2, 2), strides=2),\n",
        "    tf.keras.layers.Flatten(),\n",
        "    tf.keras.layers.Dense(128, activation=tf.nn.relu),\n",
        "    tf.keras.layers.Dense(10,  activation=tf.nn.softmax)\n",
        "])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "gut8A_7rCaW6"
      },
      "source": [
        "এখন নেটওয়ার্ক লেয়ারের গল্প:\n",
        "\n",
        "* **\"কনভল্যুশন\"** `tf.keras.layers.Conv2D এবং MaxPooling2D`— আমাদের নেটওয়ার্ক শুরু হয়েছে দুই জোড়া কনভল্যুশন/ম্যাক্সপুল এলিমেন্ট দিয়ে। প্রথম লেয়ার হচ্ছে একটা Conv2D (৩, ৩) ফিল্টার, কার্নাল বলতে পারেন। এই কার্নাল বসানো হচ্ছে ইনপুট ইমেজের ওপর। আসল যেই ছবিটা আছে সেটার রেজল্যুশনকে দরকারি প্যাডিং দিয়ে ৩২টা আউটপুট ইমেজ যা কনভল্যুশন হয়ে গেছে ইতিমধ্যে। আমাদেরকে বুঝতে হবে এই লেয়ারটা ৩২টা কনভলিউশন ইমেজ তৈরি করে যা ৩২টা ইনপুট ইমেজের সমান। এই কনভলিউশন (কনভলিউটেড) ইমেজ যা আসলে এসেছে কার্নাল/ফিল্টারের আউটপুট থেকে। আবারো বলছি - ইনপুট থেকে একটা ইমেজের আউটপুট হিসেবে ৩২টা কনভলিউশন ইমেজ তৈরি হয়েছে ইনপুটের সাইজে। এখন এই ৩২টা আউটপুট ইমেজকে ছোট করার পালা। সে জিনিসটা আমরা করছি ম্যাক্সপুলিং২ডি, MaxPooling2D (২,২ পিক্সেল গ্রিড), যার স্ট্রাইড সংখ্যা হচ্ছে ২। মানে দুটো পিক্সেল করে স্লাইড করছে ডানে অথবা বামে। \n",
        "\n",
        "\n",
        "* এরপরের লেয়ারটা হচ্ছে Conv2D (৩, ৩) ফিল্টার। বুঝতেই পারছেন এই ফিল্টারটা হচ্ছে (৩, ৩) কার্নাল, যা আগের ৩২টা ইমেজকে নিচ্ছে ইনপুট হিসেবে। এই ৩২টা ইমেজকে ইনপুট হিসেবে নিয়ে আউটপুট হিসেবে বের করে দিয়েছে ৬৪টা ইমেজ। এখন এই ৬৪টা ইমেজকে আমরা আবার ছোট করে নিয়ে আসবো ম্যাক্সপুলিং২ডি, MaxPooling2D (২,২ পিক্সেল গ্রিড) লেয়ার দিয়ে। এই ম্যাক্সপুলিং২ডি লেয়ারের গ্রিড হচ্ছে ২,২। এটার স্ট্রাইড হচ্ছে ২ তার মানে ২ পিক্সেল করে স্লাইডিং করছে। \n",
        "\n",
        "\n",
        "* এখানে একটা জিনিস বুঝতে পারছি এই কনভলিউশন এবং ম্যাক্সপুলিং এই দুটো জিনিস বারবার চেইন হিসেবে কাজ করছে। একটার আউটপুট আরেকটা ইনপুট - এভাবে একটা চেইন ইফেক্ট তৈরি করছে যতক্ষণ না আমাদের কাজ হচ্ছে। কয়বার ব্যবহার করবো? সেটা নির্ভর করছে আমাদের ট্রায়ালের ওপর। এই চেইন আরো ভালো কাজ করবে যখন আমরা গ্রেস্কেল ইমেজ থেকে রঙ্গিন ছবিতে চলে যাব। \n",
        "\n",
        "\n",
        "* **আউটপুট** `tf.keras.layers.Dense` — এই সাধারন নিউরাল নেটওয়ার্কের শুরুতেই আমরা ‘ফ্ল্যাটেনিং’ করে নিচ্ছি ইনপুটে। একদম আগের গল্প। ছবি যাই হোক না কেন সেটাকে এক লাইনে দাড়া করাচ্ছি আমাদের ইনপুটের জন্য। এরপর আমরা এখানে ১২৮টা নিউরন দিয়ে একটা ‘ডেন্স’ (পাশাপাশি কানেক্টেড) লেয়ার তৈরি করেছি। এটার অ্যাক্টিভেশন ফাংশন হচ্ছে ‘রেল্যু’, এটাই ভালো এই কাজে। \n",
        "\n",
        "\n",
        "* আর শেষ লেয়ারটা কি হবে? ঠিক ধরেছেন, আমাদের যেহেতু দশটা অবজেক্ট মানে কাপড়চোপড়, জুতা, স্নিকার ইত্যাদি ইত্যাদি জিনিসকে ঠিকমতো ক্লাসিফাই করতে হবে - সেজন্য দরকার ১০টা নোডের একটা লেয়ার। প্রতিটা নোড ‘রিপ্রেজেন্ট’ করছে একটা করে কাপড়চোপড়ের ক্লাস। আমাদের এই সর্বশেষ লেয়ার ইনপুট নিচ্ছে ১২৮টা নোড থেকে, আর সেটার আউটপুট পাঠাচ্ছে ০ আর ১ এর মধ্যে। ০ এবং ১ সংখ্যার মধ্যে প্রোবাবিলিটি ডিস্ট্রিবিউশন যেটার সবচেয়ে বেশি হবে সেদিকেই সেই ক্লাসটাকে আইডেন্টিফাই করবে। আগেও বলেছি এই দশটা নোডের যোগফল হচ্ছে ১। \n",
        "\n",
        "### মডেল কম্পাইলেশন \n",
        "\n",
        "আলাপ করেছি আগের নিউরাল নেটওয়ার্কে। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Lhan11blCaW7",
        "colab": {}
      },
      "source": [
        "model.compile(optimizer='adam',\n",
        "              loss='sparse_categorical_crossentropy',\n",
        "              metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "qKF6uW-BCaW-"
      },
      "source": [
        "## মডেলকে ট্রেইন করি \n",
        "\n",
        "এতো ডেটা, এবার আমরা একটু অন্যভাবে মডেল ট্রেইন করি। এই ব্যাপারটা আগের মডেলেও চলবে। ট্রেইন ডেটাসেটের কিছু নতুন ধারণা - ব্যাচ নিয়ে। ব্যাচে ট্রেনিং। \n",
        "\n",
        "১. BATCH_SIZE = 32 অথবা `dataset.batch(32)` মানে হচ্ছে কতগুলো ইমেজ সে প্রসেস করবে ১ ব্যাচে `model.fit` ব্যবহার করার সময়। এই ফাঁকে মডেলের ভ্যারিয়েবলগুলো আপডেট হবে। \n",
        "\n",
        "২. `dataset.repeat()` মানে এই প্রসেস বারবার হবে শেষ না হওয়া পর্যন্ত। `epochs` বলছে পুরো ডেটাসেট কতোবার পুরোটাই 'আইটারেট' করবে। \n",
        "\n",
        "৩. shuffle(num_train_examples) মানে `dataset.shuffle(60000)` পুরো ডেটাসেটকে দৈবচয়নের ভিত্তিতে এমনভাবে ঘুটিয়ে দেবে যাতে সে আর শিখতে না পারে।  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "o_Dp8971McQ1",
        "colab": {}
      },
      "source": [
        "BATCH_SIZE = 32\n",
        "train_dataset = train_dataset.repeat().shuffle(num_train_examples).batch(BATCH_SIZE)\n",
        "test_dataset = test_dataset.batch(BATCH_SIZE)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "RSTMftEqwF7l",
        "colab_type": "text"
      },
      "source": [
        "আমরা `model.fit` মেথড কল করলে এই কাজগুলো হবে:\n",
        "\n",
        "১. মডেলে ফিড হচ্ছে ট্রেনিং ডেটা, `train_dataset`।\n",
        "2. মডেল শিখছে তার ইমেজ আর সেটার লেবেল থেকে। `train_dataset` এর মধ্যে ইমেজ এবং লেবেল দুটোই আছে। \n",
        "৩. ৫ * ৬০,০০০ = ৩০,০০০০ এক্সাম্পল এর সাথে দরকারি স্টেপ।  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "xvwvpA64CaW_",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        },
        "outputId": "848f8315-e7d4-4e72-9a65-d6ff3292e2b2"
      },
      "source": [
        "model.fit(train_dataset, epochs=10, steps_per_epoch=math.ceil(num_train_examples/BATCH_SIZE))"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train for 1875 steps\n",
            "Epoch 1/10\n",
            "1875/1875 [==============================] - 54s 29ms/step - loss: 0.3978 - accuracy: 0.8572\n",
            "Epoch 2/10\n",
            "1875/1875 [==============================] - 34s 18ms/step - loss: 0.2560 - accuracy: 0.9072\n",
            "Epoch 3/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.2149 - accuracy: 0.9214\n",
            "Epoch 4/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.1817 - accuracy: 0.9329\n",
            "Epoch 5/10\n",
            "1875/1875 [==============================] - 31s 17ms/step - loss: 0.1566 - accuracy: 0.9422\n",
            "Epoch 6/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.1367 - accuracy: 0.9495\n",
            "Epoch 7/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.1110 - accuracy: 0.9588\n",
            "Epoch 8/10\n",
            "1875/1875 [==============================] - 31s 17ms/step - loss: 0.0943 - accuracy: 0.9652\n",
            "Epoch 9/10\n",
            "1875/1875 [==============================] - 32s 17ms/step - loss: 0.0786 - accuracy: 0.9713\n",
            "Epoch 10/10\n",
            "1875/1875 [==============================] - 31s 16ms/step - loss: 0.0636 - accuracy: 0.9763\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<tensorflow.python.keras.callbacks.History at 0x7fd8702f7d30>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 25
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "W3ZVOhugCaXA"
      },
      "source": [
        "মডেল ট্রেনিং এর সময় সেটার লস এবং অ্যাক্যুরেসি ম্যাট্রিক্স দেখলে বোঝা যায় ট্রেনিং ডেটার ওপর মডেল ভালো করেছে। এটাই স্বাভাবিক। "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "oEw4bZgGCaXB"
      },
      "source": [
        "## এখন আমরা অ্যাক্যুরেসি বের করবো কিভাবে?\n",
        "\n",
        "অবশ্যই টেস্ট ডেটাসেট দিয়ে দিয়ে। তার সবগুলো এক্সাম্পল দিয়ে। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "VflXLEeECaXC",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "f43bdaf4-5b49-4423-dbd9-d85f4939e007"
      },
      "source": [
        "test_loss, test_accuracy = model.evaluate(test_dataset, steps=math.ceil(num_test_examples/32))\n",
        "print('Accuracy on test dataset:', test_accuracy)"
      ],
      "execution_count": 26,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "313/313 [==============================] - 5s 14ms/step - loss: 0.3280 - accuracy: 0.9157\n",
            "Accuracy on test dataset: 0.9157\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "yWfgsmVXCaXG"
      },
      "source": [
        "সাধারণতঃ `train_dataset` থেকে `test_dataset` এর অ্যাক্যুরেসি কিছুটা কম হয়। এটাই স্বাভাবিক। কারণে সেতো ট্রেনিং করেছে `train_dataset` এর ওপর। এদিকে `test_dataset` তার কাছে অজানা ডেটা। এটাকে ওভারফিটিং বলে। "
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "xsoS7CPDCaXH"
      },
      "source": [
        "## মডেল থেকে কিছু প্রেডিকশন করি "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Ccoz4conNCpl",
        "colab": {}
      },
      "source": [
        "for test_images, test_labels in test_dataset.take(1):\n",
        "  test_images = test_images.numpy()\n",
        "  test_labels = test_labels.numpy()\n",
        "  predictions = model.predict(test_images)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Gl91RPhdCaXI",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "f458918a-60b7-4894-d783-8d3fd47fd030"
      },
      "source": [
        "predictions.shape"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(32, 10)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 28
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "x9Kk1voUCaXJ"
      },
      "source": [
        "টেস্টসেট থেকে মডেল কিছু প্রেডিকশন করেছে। শুরুর ইমেজটা দেখি। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "3DmJEUinCaXK",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 69
        },
        "outputId": "6739c4e4-953f-41f7-8947-c5c541f2587b"
      },
      "source": [
        "predictions[0]"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([3.0347110e-08, 9.7811360e-12, 6.1781280e-04, 3.9071399e-08,\n",
              "       3.6942250e-01, 3.2189834e-10, 6.2995964e-01, 5.8712570e-11,\n",
              "       7.5262943e-09, 2.7785850e-11], dtype=float32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "-hw1hgeSCaXN"
      },
      "source": [
        "আমরা জানি যে আমাদের প্রেডিকশন অ্যারে কিন্তু ১০টা সংখ্যায় আসবে। কারণ আমাদের আউটপুট অবজেক্ট হচ্ছে ১০টা। তবে ১০টার আউটপুট তার 'করেসপন্ডিং' কনফিডেন্স লেভেল, কাপড়চোপড়, স্নিকার, জুতা ইত্যাদির জন্য। আমাদেরকে বের করতে হবে সবচেয়ে বেশি কনফিডেন্ট কোথায়? সবচেয়ে বড় ভ্যালুতে।  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "qsqenuPnCaXO",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "96e2c453-5768-438e-bc8e-cda3f9d5710e"
      },
      "source": [
        "np.argmax(predictions[0])"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "6"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "E51yS7iCCaXO"
      },
      "source": [
        "আমরা বুঝতে পারছি, মডেল কনফিডেন্ট একটা শার্টের ব্যাপারে মানে `class_names[6]`। তাহলে আমাদের টেস্ট লেবেলটা দেখি ছবিতে। "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Sd7Pgsu6CaXP",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "outputId": "5018165a-03e0-4b88-8e7a-fe488c9b1ee5"
      },
      "source": [
        "test_labels[0]"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "6"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 31
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "ygh2yYC972ne"
      },
      "source": [
        "## মডেল সামারি দেখি  "
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Qo_e3CS_zg2V",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 399
        },
        "outputId": "d9765769-973a-461f-811f-0bbbb52ba568"
      },
      "source": [
        "# pd.DataFrame(model.history).plot(ylim=(-0.05, 1.05))\n",
        "model.summary()"
      ],
      "execution_count": 45,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "conv2d (Conv2D)              (None, 28, 28, 32)        320       \n",
            "_________________________________________________________________\n",
            "max_pooling2d (MaxPooling2D) (None, 14, 14, 32)        0         \n",
            "_________________________________________________________________\n",
            "conv2d_1 (Conv2D)            (None, 14, 14, 64)        18496     \n",
            "_________________________________________________________________\n",
            "max_pooling2d_1 (MaxPooling2 (None, 7, 7, 64)          0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 3136)              0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 128)               401536    \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 10)                1290      \n",
            "=================================================================\n",
            "Total params: 421,642\n",
            "Trainable params: 421,642\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}