gann.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Import Dependencies and Data**\n",
    "\n",
    "In this section, we are laying the foundation for our Generative Adversarial Neural Network (GANN) that will attempt to generate images of fashion items. Specifically, we will use TensorFlow and TensorFlow Datasets (TFDS) to load the fashion_mnist dataset. The dataset consists of grayscale images of various clothing items, and we will train two models:\n",
    "\n",
    "* Generator Model: This model will generate new images that mimic real fashion items.\n",
    "* Discriminator Model: This model will learn to differentiate between the real images from the dataset and the fake ones generated by the generator. Together, these two models form a GANN that improves through adversarial training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: tensorflow in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (2.17.0)\n",
      "Requirement already satisfied: matplotlib in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (3.8.2)\n",
      "Requirement already satisfied: tensorflow-datasets in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (4.9.6)\n",
      "Requirement already satisfied: ipywidgets in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (8.1.1)\n",
      "Requirement already satisfied: tensorflow-intel==2.17.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow) (2.17.0)\n",
      "Requirement already satisfied: absl-py>=1.0.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (2.0.0)\n",
      "Requirement already satisfied: astunparse>=1.6.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (1.6.3)\n",
      "Requirement already satisfied: flatbuffers>=24.3.25 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (24.3.25)\n",
      "Requirement already satisfied: gast!=0.5.0,!=0.5.1,!=0.5.2,>=0.2.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (0.4.0)\n",
      "Requirement already satisfied: google-pasta>=0.1.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (0.2.0)\n",
      "Requirement already satisfied: h5py>=3.10.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (3.10.0)\n",
      "Requirement already satisfied: libclang>=13.0.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (16.0.6)\n",
      "Requirement already satisfied: ml-dtypes<0.5.0,>=0.3.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (0.4.1)\n",
      "Requirement already satisfied: opt-einsum>=2.3.2 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (3.3.0)\n",
      "Requirement already satisfied: packaging in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (23.2)\n",
      "Requirement already satisfied: protobuf!=4.21.0,!=4.21.1,!=4.21.2,!=4.21.3,!=4.21.4,!=4.21.5,<5.0.0dev,>=3.20.3 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (4.25.5)\n",
      "Requirement already satisfied: requests<3,>=2.21.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (2.31.0)\n",
      "Requirement already satisfied: setuptools in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (75.1.0)\n",
      "Requirement already satisfied: six>=1.12.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (1.16.0)\n",
      "Requirement already satisfied: termcolor>=1.1.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (2.4.0)\n",
      "Requirement already satisfied: typing-extensions>=3.6.6 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (4.5.0)\n",
      "Requirement already satisfied: wrapt>=1.11.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (1.14.1)\n",
      "Requirement already satisfied: grpcio<2.0,>=1.24.3 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (1.59.3)\n",
      "Requirement already satisfied: tensorboard<2.18,>=2.17 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (2.17.1)\n",
      "Requirement already satisfied: keras>=3.2.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (3.6.0)\n",
      "Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (0.31.0)\n",
      "Requirement already satisfied: numpy<2.0.0,>=1.23.5 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-intel==2.17.0->tensorflow) (1.23.5)\n",
      "Requirement already satisfied: contourpy>=1.0.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from matplotlib) (1.2.0)\n",
      "Requirement already satisfied: cycler>=0.10 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from matplotlib) (0.12.1)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from matplotlib) (4.46.0)\n",
      "Requirement already satisfied: kiwisolver>=1.3.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from matplotlib) (1.4.5)\n",
      "Requirement already satisfied: pillow>=8 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from matplotlib) (9.5.0)\n",
      "Requirement already satisfied: pyparsing>=2.3.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from matplotlib) (3.1.1)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from matplotlib) (2.8.2)\n",
      "Requirement already satisfied: click in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (8.1.7)\n",
      "Requirement already satisfied: dm-tree in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (0.1.8)\n",
      "Requirement already satisfied: immutabledict in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (4.2.0)\n",
      "Requirement already satisfied: promise in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (2.3)\n",
      "Requirement already satisfied: psutil in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (5.9.6)\n",
      "Requirement already satisfied: pyarrow in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (17.0.0)\n",
      "Requirement already satisfied: simple-parsing in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (0.1.6)\n",
      "Requirement already satisfied: tensorflow-metadata in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (1.16.1)\n",
      "Requirement already satisfied: toml in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (0.10.2)\n",
      "Requirement already satisfied: tqdm in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-datasets) (4.66.5)\n",
      "Requirement already satisfied: etils>=1.9.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from etils[enp,epath,epy,etree]>=1.9.1; python_version >= \"3.11\"->tensorflow-datasets) (1.9.4)\n",
      "Requirement already satisfied: comm>=0.1.3 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipywidgets) (0.2.0)\n",
      "Requirement already satisfied: ipython>=6.1.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipywidgets) (8.18.1)\n",
      "Requirement already satisfied: traitlets>=4.3.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipywidgets) (5.14.0)\n",
      "Requirement already satisfied: widgetsnbextension~=4.0.9 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipywidgets) (4.0.9)\n",
      "Requirement already satisfied: jupyterlab-widgets~=3.0.9 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipywidgets) (3.0.9)\n",
      "Requirement already satisfied: fsspec in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from etils[enp,epath,epy,etree]>=1.9.1; python_version >= \"3.11\"->tensorflow-datasets) (2024.9.0)\n",
      "Requirement already satisfied: importlib_resources in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from etils[enp,epath,epy,etree]>=1.9.1; python_version >= \"3.11\"->tensorflow-datasets) (6.4.5)\n",
      "Requirement already satisfied: zipp in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from etils[enp,epath,epy,etree]>=1.9.1; python_version >= \"3.11\"->tensorflow-datasets) (3.20.2)\n",
      "Requirement already satisfied: decorator in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipython>=6.1.0->ipywidgets) (5.1.1)\n",
      "Requirement already satisfied: jedi>=0.16 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipython>=6.1.0->ipywidgets) (0.19.1)\n",
      "Requirement already satisfied: matplotlib-inline in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipython>=6.1.0->ipywidgets) (0.1.6)\n",
      "Requirement already satisfied: prompt-toolkit<3.1.0,>=3.0.41 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipython>=6.1.0->ipywidgets) (3.0.41)\n",
      "Requirement already satisfied: pygments>=2.4.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipython>=6.1.0->ipywidgets) (2.17.2)\n",
      "Requirement already satisfied: stack-data in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipython>=6.1.0->ipywidgets) (0.6.3)\n",
      "Requirement already satisfied: colorama in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from ipython>=6.1.0->ipywidgets) (0.4.6)\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.17.0->tensorflow) (3.3.2)\n",
      "Requirement already satisfied: idna<4,>=2.5 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.17.0->tensorflow) (3.6)\n",
      "Requirement already satisfied: urllib3<3,>=1.21.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.17.0->tensorflow) (2.1.0)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.17.0->tensorflow) (2023.11.17)\n",
      "Requirement already satisfied: docstring-parser<1.0,>=0.15 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from simple-parsing->tensorflow-datasets) (0.16)\n",
      "Requirement already satisfied: googleapis-common-protos<2,>=1.56.4 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorflow-metadata->tensorflow-datasets) (1.65.0)\n",
      "Requirement already satisfied: wheel<1.0,>=0.23.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from astunparse>=1.6.0->tensorflow-intel==2.17.0->tensorflow) (0.44.0)\n",
      "Requirement already satisfied: parso<0.9.0,>=0.8.3 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from jedi>=0.16->ipython>=6.1.0->ipywidgets) (0.8.3)\n",
      "Requirement already satisfied: rich in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from keras>=3.2.0->tensorflow-intel==2.17.0->tensorflow) (13.7.0)\n",
      "Requirement already satisfied: namex in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from keras>=3.2.0->tensorflow-intel==2.17.0->tensorflow) (0.0.7)\n",
      "Requirement already satisfied: optree in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from keras>=3.2.0->tensorflow-intel==2.17.0->tensorflow) (0.13.0)\n",
      "Requirement already satisfied: wcwidth in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from prompt-toolkit<3.1.0,>=3.0.41->ipython>=6.1.0->ipywidgets) (0.2.12)\n",
      "Requirement already satisfied: markdown>=2.6.8 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorboard<2.18,>=2.17->tensorflow-intel==2.17.0->tensorflow) (3.5.1)\n",
      "Requirement already satisfied: tensorboard-data-server<0.8.0,>=0.7.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorboard<2.18,>=2.17->tensorflow-intel==2.17.0->tensorflow) (0.7.2)\n",
      "Requirement already satisfied: werkzeug>=1.0.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from tensorboard<2.18,>=2.17->tensorflow-intel==2.17.0->tensorflow) (3.0.1)\n",
      "Requirement already satisfied: executing>=1.2.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from stack-data->ipython>=6.1.0->ipywidgets) (2.0.1)\n",
      "Requirement already satisfied: asttokens>=2.1.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from stack-data->ipython>=6.1.0->ipywidgets) (2.4.1)\n",
      "Requirement already satisfied: pure-eval in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from stack-data->ipython>=6.1.0->ipywidgets) (0.2.2)\n",
      "Requirement already satisfied: MarkupSafe>=2.1.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from werkzeug>=1.0.1->tensorboard<2.18,>=2.17->tensorflow-intel==2.17.0->tensorflow) (2.1.3)\n",
      "Requirement already satisfied: markdown-it-py>=2.2.0 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from rich->keras>=3.2.0->tensorflow-intel==2.17.0->tensorflow) (3.0.0)\n",
      "Requirement already satisfied: mdurl~=0.1 in c:\\users\\amir\\appdata\\local\\packages\\pythonsoftwarefoundation.python.3.11_qbz5n2kfra8p0\\localcache\\local-packages\\python311\\site-packages (from markdown-it-py>=2.2.0->rich->keras>=3.2.0->tensorflow-intel==2.17.0->tensorflow) (0.1.2)\n",
      "Note: you may need to restart the kernel to use updated packages.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "[notice] A new release of pip is available: 23.3.2 -> 24.2\n",
      "[notice] To update, run: C:\\Users\\Amir\\AppData\\Local\\Microsoft\\WindowsApps\\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\\python.exe -m pip install --upgrade pip\n"
     ]
    }
   ],
   "source": [
    "#installing all the dependencies\n",
    "%pip install tensorflow matplotlib tensorflow-datasets ipywidgets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Bringing in tensorflow\n",
    "import tensorflow as tf\n",
    "gpus = tf.config.experimental.list_physical_devices('GPU')\n",
    "for gpu in gpus: \n",
    "    tf.config.experimental.set_memory_growth(gpu[0], True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Brining in tensorflow datasets for fashion mnist \n",
    "import tensorflow_datasets as tfds\n",
    "# Bringing in matplotlib for viz stuff\n",
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Use the tensorflow datasets api to bring in the data source\n",
    "ds = tfds.load('fashion_mnist', split='train')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<_PrefetchDataset element_spec={'image': TensorSpec(shape=(28, 28, 1), dtype=tf.uint8, name=None), 'label': TensorSpec(shape=(), dtype=tf.int64, name=None)}>\n",
      "<class 'tensorflow.python.data.ops.prefetch_op._PrefetchDataset'>\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "dict_keys(['image', 'label'])"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#exploring the ds\n",
    "print(ds)\n",
    "print (type(ds))\n",
    "ds.as_numpy_iterator().next().keys()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**visualizing Data and Build Dataset**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 18],\n",
       "        [ 77],\n",
       "        [227],\n",
       "        [227],\n",
       "        [208],\n",
       "        [210],\n",
       "        [225],\n",
       "        [216],\n",
       "        [ 85],\n",
       "        [ 32],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 61],\n",
       "        [100],\n",
       "        [ 97],\n",
       "        [ 80],\n",
       "        [ 57],\n",
       "        [117],\n",
       "        [227],\n",
       "        [238],\n",
       "        [115],\n",
       "        [ 49],\n",
       "        [ 78],\n",
       "        [106],\n",
       "        [108],\n",
       "        [ 71],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 81],\n",
       "        [105],\n",
       "        [ 80],\n",
       "        [ 69],\n",
       "        [ 72],\n",
       "        [ 64],\n",
       "        [ 44],\n",
       "        [ 21],\n",
       "        [ 13],\n",
       "        [ 44],\n",
       "        [ 69],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 80],\n",
       "        [114],\n",
       "        [ 80],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 26],\n",
       "        [ 92],\n",
       "        [ 69],\n",
       "        [ 68],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 71],\n",
       "        [ 74],\n",
       "        [ 83],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 78],\n",
       "        [ 74],\n",
       "        [ 74],\n",
       "        [ 83],\n",
       "        [ 77],\n",
       "        [108],\n",
       "        [ 34],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 55],\n",
       "        [ 92],\n",
       "        [ 69],\n",
       "        [ 74],\n",
       "        [ 74],\n",
       "        [ 71],\n",
       "        [ 71],\n",
       "        [ 77],\n",
       "        [ 69],\n",
       "        [ 66],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 77],\n",
       "        [ 80],\n",
       "        [ 80],\n",
       "        [ 78],\n",
       "        [ 94],\n",
       "        [ 63],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 63],\n",
       "        [ 95],\n",
       "        [ 66],\n",
       "        [ 68],\n",
       "        [ 72],\n",
       "        [ 72],\n",
       "        [ 69],\n",
       "        [ 72],\n",
       "        [ 74],\n",
       "        [ 74],\n",
       "        [ 74],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 80],\n",
       "        [ 77],\n",
       "        [106],\n",
       "        [ 61],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 80],\n",
       "        [108],\n",
       "        [ 71],\n",
       "        [ 69],\n",
       "        [ 72],\n",
       "        [ 71],\n",
       "        [ 69],\n",
       "        [ 72],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 72],\n",
       "        [ 72],\n",
       "        [ 75],\n",
       "        [ 78],\n",
       "        [ 72],\n",
       "        [ 85],\n",
       "        [128],\n",
       "        [ 64],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 88],\n",
       "        [120],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 72],\n",
       "        [ 77],\n",
       "        [ 74],\n",
       "        [ 74],\n",
       "        [ 77],\n",
       "        [ 78],\n",
       "        [ 83],\n",
       "        [ 83],\n",
       "        [ 66],\n",
       "        [111],\n",
       "        [123],\n",
       "        [ 78],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 85],\n",
       "        [134],\n",
       "        [ 74],\n",
       "        [ 85],\n",
       "        [ 69],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 81],\n",
       "        [ 75],\n",
       "        [ 61],\n",
       "        [151],\n",
       "        [115],\n",
       "        [ 91],\n",
       "        [ 12],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 10],\n",
       "        [ 85],\n",
       "        [153],\n",
       "        [ 83],\n",
       "        [ 80],\n",
       "        [ 68],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 80],\n",
       "        [ 68],\n",
       "        [ 61],\n",
       "        [162],\n",
       "        [122],\n",
       "        [ 78],\n",
       "        [  6],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 30],\n",
       "        [ 75],\n",
       "        [154],\n",
       "        [ 85],\n",
       "        [ 80],\n",
       "        [ 71],\n",
       "        [ 80],\n",
       "        [ 72],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 78],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 49],\n",
       "        [191],\n",
       "        [132],\n",
       "        [ 72],\n",
       "        [ 15],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 58],\n",
       "        [ 66],\n",
       "        [174],\n",
       "        [115],\n",
       "        [ 66],\n",
       "        [ 77],\n",
       "        [ 80],\n",
       "        [ 72],\n",
       "        [ 78],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 78],\n",
       "        [ 78],\n",
       "        [ 77],\n",
       "        [ 66],\n",
       "        [ 49],\n",
       "        [222],\n",
       "        [131],\n",
       "        [ 77],\n",
       "        [ 37],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 69],\n",
       "        [ 55],\n",
       "        [179],\n",
       "        [139],\n",
       "        [ 55],\n",
       "        [ 92],\n",
       "        [ 74],\n",
       "        [ 74],\n",
       "        [ 78],\n",
       "        [ 74],\n",
       "        [ 78],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 80],\n",
       "        [ 64],\n",
       "        [ 55],\n",
       "        [242],\n",
       "        [111],\n",
       "        [ 95],\n",
       "        [ 44],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 74],\n",
       "        [ 57],\n",
       "        [159],\n",
       "        [180],\n",
       "        [ 55],\n",
       "        [ 92],\n",
       "        [ 64],\n",
       "        [ 72],\n",
       "        [ 74],\n",
       "        [ 74],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 78],\n",
       "        [ 55],\n",
       "        [ 66],\n",
       "        [255],\n",
       "        [ 97],\n",
       "        [108],\n",
       "        [ 49],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 74],\n",
       "        [ 66],\n",
       "        [145],\n",
       "        [153],\n",
       "        [ 72],\n",
       "        [ 83],\n",
       "        [ 58],\n",
       "        [ 78],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 72],\n",
       "        [ 80],\n",
       "        [ 30],\n",
       "        [132],\n",
       "        [255],\n",
       "        [ 37],\n",
       "        [122],\n",
       "        [ 60],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 80],\n",
       "        [ 69],\n",
       "        [142],\n",
       "        [180],\n",
       "        [142],\n",
       "        [ 57],\n",
       "        [ 64],\n",
       "        [ 78],\n",
       "        [ 74],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 72],\n",
       "        [ 85],\n",
       "        [ 21],\n",
       "        [185],\n",
       "        [227],\n",
       "        [ 37],\n",
       "        [143],\n",
       "        [ 63],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 83],\n",
       "        [ 71],\n",
       "        [136],\n",
       "        [194],\n",
       "        [126],\n",
       "        [ 46],\n",
       "        [ 69],\n",
       "        [ 75],\n",
       "        [ 72],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 78],\n",
       "        [ 38],\n",
       "        [139],\n",
       "        [185],\n",
       "        [ 60],\n",
       "        [151],\n",
       "        [ 58],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  4],\n",
       "        [ 81],\n",
       "        [ 74],\n",
       "        [145],\n",
       "        [177],\n",
       "        [ 78],\n",
       "        [ 49],\n",
       "        [ 74],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 72],\n",
       "        [ 63],\n",
       "        [ 80],\n",
       "        [156],\n",
       "        [117],\n",
       "        [153],\n",
       "        [ 55],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 10],\n",
       "        [ 80],\n",
       "        [ 72],\n",
       "        [157],\n",
       "        [163],\n",
       "        [ 61],\n",
       "        [ 55],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 71],\n",
       "        [ 60],\n",
       "        [ 98],\n",
       "        [156],\n",
       "        [132],\n",
       "        [ 58],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 13],\n",
       "        [ 77],\n",
       "        [ 74],\n",
       "        [157],\n",
       "        [143],\n",
       "        [ 43],\n",
       "        [ 61],\n",
       "        [ 72],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 77],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 71],\n",
       "        [ 58],\n",
       "        [ 80],\n",
       "        [157],\n",
       "        [120],\n",
       "        [ 66],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 18],\n",
       "        [ 81],\n",
       "        [ 74],\n",
       "        [156],\n",
       "        [114],\n",
       "        [ 35],\n",
       "        [ 72],\n",
       "        [ 71],\n",
       "        [ 75],\n",
       "        [ 78],\n",
       "        [ 72],\n",
       "        [ 66],\n",
       "        [ 80],\n",
       "        [ 78],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 64],\n",
       "        [ 63],\n",
       "        [165],\n",
       "        [119],\n",
       "        [ 68],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 23],\n",
       "        [ 85],\n",
       "        [ 81],\n",
       "        [177],\n",
       "        [ 57],\n",
       "        [ 52],\n",
       "        [ 77],\n",
       "        [ 71],\n",
       "        [ 78],\n",
       "        [ 80],\n",
       "        [ 72],\n",
       "        [ 75],\n",
       "        [ 74],\n",
       "        [ 77],\n",
       "        [ 77],\n",
       "        [ 75],\n",
       "        [ 64],\n",
       "        [ 37],\n",
       "        [173],\n",
       "        [ 95],\n",
       "        [ 72],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 26],\n",
       "        [ 81],\n",
       "        [ 86],\n",
       "        [160],\n",
       "        [ 20],\n",
       "        [ 75],\n",
       "        [ 77],\n",
       "        [ 77],\n",
       "        [ 80],\n",
       "        [ 78],\n",
       "        [ 80],\n",
       "        [ 89],\n",
       "        [ 78],\n",
       "        [ 81],\n",
       "        [ 83],\n",
       "        [ 80],\n",
       "        [ 74],\n",
       "        [ 20],\n",
       "        [177],\n",
       "        [ 77],\n",
       "        [ 74],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 49],\n",
       "        [ 77],\n",
       "        [ 91],\n",
       "        [200],\n",
       "        [  0],\n",
       "        [ 83],\n",
       "        [ 95],\n",
       "        [ 86],\n",
       "        [ 88],\n",
       "        [ 88],\n",
       "        [ 89],\n",
       "        [ 88],\n",
       "        [ 89],\n",
       "        [ 88],\n",
       "        [ 83],\n",
       "        [ 89],\n",
       "        [ 86],\n",
       "        [  0],\n",
       "        [191],\n",
       "        [ 78],\n",
       "        [ 80],\n",
       "        [ 24],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 54],\n",
       "        [ 71],\n",
       "        [108],\n",
       "        [165],\n",
       "        [  0],\n",
       "        [ 24],\n",
       "        [ 57],\n",
       "        [ 52],\n",
       "        [ 57],\n",
       "        [ 60],\n",
       "        [ 60],\n",
       "        [ 60],\n",
       "        [ 63],\n",
       "        [ 63],\n",
       "        [ 77],\n",
       "        [ 89],\n",
       "        [ 52],\n",
       "        [  0],\n",
       "        [211],\n",
       "        [ 97],\n",
       "        [ 77],\n",
       "        [ 61],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 68],\n",
       "        [ 91],\n",
       "        [117],\n",
       "        [137],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 18],\n",
       "        [216],\n",
       "        [ 94],\n",
       "        [ 97],\n",
       "        [ 57],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 54],\n",
       "        [115],\n",
       "        [105],\n",
       "        [185],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  1],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [153],\n",
       "        [ 78],\n",
       "        [106],\n",
       "        [ 37],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]],\n",
       "\n",
       "       [[  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [ 18],\n",
       "        [ 61],\n",
       "        [ 41],\n",
       "        [103],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [106],\n",
       "        [ 47],\n",
       "        [ 69],\n",
       "        [ 23],\n",
       "        [  0],\n",
       "        [  0],\n",
       "        [  0]]], dtype=uint8)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Do some data transformation\n",
    "import numpy as np\n",
    "\n",
    "# Setup connection aka iterator\n",
    "dataiterator = ds.as_numpy_iterator()\n",
    "# Getting data out of the pipeline\n",
    "dataiterator.next()['image']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x2000 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Setup the subplot formatting \n",
    "fig, ax = plt.subplots(ncols=4, figsize=(20,20))\n",
    "# Loop four times and get images \n",
    "for idx in range(4): \n",
    "    # Grab an image and label\n",
    "    sample = dataiterator.next()\n",
    "    # Plot the image using a specific subplot \n",
    "    ax[idx].imshow(np.squeeze(sample['image']))\n",
    "    # Appending the image label as the plot title \n",
    "    ax[idx].title.set_text(sample['label'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scale and return images only \n",
    "def scale_images(data): \n",
    "    image = data['image']\n",
    "    return image / 255\n",
    "\n",
    "# Reload the dataset \n",
    "ds = tfds.load('fashion_mnist', split='train')\n",
    "# Running the dataset through the scale_images preprocessing step\n",
    "ds = ds.map(scale_images) \n",
    "# Cache the dataset for that batch \n",
    "ds = ds.cache()\n",
    "# Shuffle it up \n",
    "ds = ds.shuffle(60000)\n",
    "# Batch into 128 images per sample\n",
    "ds = ds.batch(128)\n",
    "# Reduces the likelihood of bottleneck \n",
    "ds = ds.prefetch(64)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(128, 28, 28, 1)"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#getting the finalized input shape\n",
    "ds.as_numpy_iterator().next().shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Building our Neural Network**\n",
    "\n",
    "In this section, we focus on visualizing the images from the Fashion MNIST dataset and preparing the dataset for training our GANN model. We'll:\n",
    "\n",
    "* Visualize the data: By plotting a few sample images from the dataset, we get a visual understanding of what the clothing items look like.\n",
    "* Build the dataset pipeline: We'll preprocess the data by normalizing the pixel values and batching it for efficient training. To avoid bottlenecks during model training, we'll also shuffle and prefetch the data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Bring in the sequential api for the generator and discriminator\n",
    "from tensorflow.keras.models import Sequential\n",
    "# Bring in the layers for the neural network\n",
    "from tensorflow.keras.layers import Conv2D, Dense, Flatten, Reshape, LeakyReLU, Dropout, UpSampling2D"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_generator(): \n",
    "    model = Sequential()\n",
    "    \n",
    "    # Takes in random values and reshapes it to 7x7x128\n",
    "    # Beginnings of a generated image\n",
    "    model.add(Dense(7*7*128, input_dim=128))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    model.add(Reshape((7,7,128)))\n",
    "    \n",
    "    # Upsampling block 1 \n",
    "    model.add(UpSampling2D())\n",
    "    model.add(Conv2D(128, 5, padding='same'))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    \n",
    "    # Upsampling block 2 \n",
    "    model.add(UpSampling2D())\n",
    "    model.add(Conv2D(128, 5, padding='same'))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    \n",
    "    # Convolutional block 1\n",
    "    model.add(Conv2D(128, 4, padding='same'))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    \n",
    "    # Convolutional block 2\n",
    "    model.add(Conv2D(128, 4, padding='same'))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    \n",
    "    # Conv layer to get to one channel\n",
    "    model.add(Conv2D(1, 4, padding='same', activation='sigmoid'))\n",
    "    \n",
    "    return model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Amir\\AppData\\Local\\Packages\\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\\LocalCache\\local-packages\\Python311\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "generator = build_generator()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential\"</span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1mModel: \"sequential\"\u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ dense (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                   │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">6272</span>)           │       <span style=\"color: #00af00; text-decoration-color: #00af00\">809,088</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)         │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">6272</span>)           │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ reshape (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Reshape</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">7</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)      │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ up_sampling2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">UpSampling2D</span>)    │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">14</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">14</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">14</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">14</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │       <span style=\"color: #00af00; text-decoration-color: #00af00\">409,728</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">14</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">14</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ up_sampling2d_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">UpSampling2D</span>)  │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │       <span style=\"color: #00af00; text-decoration-color: #00af00\">409,728</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │       <span style=\"color: #00af00; text-decoration-color: #00af00\">262,272</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │       <span style=\"color: #00af00; text-decoration-color: #00af00\">262,272</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_4 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_4 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">28</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1</span>)      │         <span style=\"color: #00af00; text-decoration-color: #00af00\">2,049</span> │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "</pre>\n"
      ],
      "text/plain": [
       "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ dense (\u001b[38;5;33mDense\u001b[0m)                   │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m6272\u001b[0m)           │       \u001b[38;5;34m809,088\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu (\u001b[38;5;33mLeakyReLU\u001b[0m)         │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m6272\u001b[0m)           │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ reshape (\u001b[38;5;33mReshape\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m7\u001b[0m, \u001b[38;5;34m128\u001b[0m)      │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ up_sampling2d (\u001b[38;5;33mUpSampling2D\u001b[0m)    │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m14\u001b[0m, \u001b[38;5;34m14\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d (\u001b[38;5;33mConv2D\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m14\u001b[0m, \u001b[38;5;34m14\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │       \u001b[38;5;34m409,728\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_1 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m14\u001b[0m, \u001b[38;5;34m14\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ up_sampling2d_1 (\u001b[38;5;33mUpSampling2D\u001b[0m)  │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_1 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │       \u001b[38;5;34m409,728\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_2 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_2 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │       \u001b[38;5;34m262,272\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_3 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_3 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │       \u001b[38;5;34m262,272\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_4 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_4 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m28\u001b[0m, \u001b[38;5;34m1\u001b[0m)      │         \u001b[38;5;34m2,049\u001b[0m │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">2,155,137</span> (8.22 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m2,155,137\u001b[0m (8.22 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">2,155,137</span> (8.22 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m2,155,137\u001b[0m (8.22 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "generator.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 25ms/step\n"
     ]
    }
   ],
   "source": [
    "img = generator.predict(np.random.randn(4,128,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(4, 28, 28, 1)"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 20ms/step\n",
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 40ms/step\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 2000x2000 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "img = generator.predict(np.random.randn(4,128,1))\n",
    "\n",
    "# Generate new fashion\n",
    "img = generator.predict(np.random.randn(4,128,1))\n",
    "# Setup the subplot formatting \n",
    "fig, ax = plt.subplots(ncols=4, figsize=(20,20))\n",
    "# Loop four times and get images \n",
    "for idx, img in enumerate(img): \n",
    "    # Plot the image using a specific subplot \n",
    "    ax[idx].imshow(np.squeeze(img))\n",
    "    # Appending the image label as the plot title \n",
    "    ax[idx].title.set_text(idx)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Build Discriminator**\n",
    "\n",
    "In this section, we will construct the Discriminator model for our GANN. The Discriminator's role is to distinguish between real images from the Fashion MNIST dataset and fake images generated by the Generator model. By using a series of convolutional layers with Leaky ReLU activations, the Discriminator will learn to identify features in the images that indicate authenticity. The model concludes with a Dense layer that outputs a probability score indicating whether the input image is real or fake."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "def build_discriminator(): \n",
    "    model = Sequential()\n",
    "    \n",
    "    # First Conv Block\n",
    "    model.add(Conv2D(32, 5, input_shape = (28,28,1)))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    model.add(Dropout(0.4))\n",
    "    \n",
    "    # Second Conv Block\n",
    "    model.add(Conv2D(64, 5))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    model.add(Dropout(0.4))\n",
    "    \n",
    "    # Third Conv Block\n",
    "    model.add(Conv2D(128, 5))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    model.add(Dropout(0.4))\n",
    "    \n",
    "    # Fourth Conv Block\n",
    "    model.add(Conv2D(256, 5))\n",
    "    model.add(LeakyReLU(0.2))\n",
    "    model.add(Dropout(0.4))\n",
    "    \n",
    "    # Flatten then pass to dense layer\n",
    "    model.add(Flatten())\n",
    "    model.add(Dropout(0.4))\n",
    "    model.add(Dense(1, activation='sigmoid'))\n",
    "    \n",
    "    return model "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\Amir\\AppData\\Local\\Packages\\PythonSoftwareFoundation.Python.3.11_qbz5n2kfra8p0\\LocalCache\\local-packages\\Python311\\site-packages\\keras\\src\\layers\\convolutional\\base_conv.py:107: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n",
      "  super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n"
     ]
    }
   ],
   "source": [
    "discriminator = build_discriminator()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\">Model: \"sequential_1\"</span>\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1mModel: \"sequential_1\"\u001b[0m\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\">┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃<span style=\"font-weight: bold\"> Layer (type)                    </span>┃<span style=\"font-weight: bold\"> Output Shape           </span>┃<span style=\"font-weight: bold\">       Param # </span>┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ conv2d_5 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">32</span>)     │           <span style=\"color: #00af00; text-decoration-color: #00af00\">832</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_5 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">32</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">24</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">32</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_6 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)     │        <span style=\"color: #00af00; text-decoration-color: #00af00\">51,264</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_6 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">20</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">64</span>)     │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_7 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">16</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">16</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │       <span style=\"color: #00af00; text-decoration-color: #00af00\">204,928</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_7 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">16</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">16</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_2 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">16</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">16</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">128</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_8 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Conv2D</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)    │       <span style=\"color: #00af00; text-decoration-color: #00af00\">819,456</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_8 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">LeakyReLU</span>)       │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_3 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">12</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">256</span>)    │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ flatten (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Flatten</span>)               │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">36864</span>)          │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_4 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dropout</span>)             │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">36864</span>)          │             <span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_1 (<span style=\"color: #0087ff; text-decoration-color: #0087ff\">Dense</span>)                 │ (<span style=\"color: #00d7ff; text-decoration-color: #00d7ff\">None</span>, <span style=\"color: #00af00; text-decoration-color: #00af00\">1</span>)              │        <span style=\"color: #00af00; text-decoration-color: #00af00\">36,865</span> │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n",
       "</pre>\n"
      ],
      "text/plain": [
       "┏━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━━━━━━━━━━┳━━━━━━━━━━━━━━━┓\n",
       "┃\u001b[1m \u001b[0m\u001b[1mLayer (type)                   \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1mOutput Shape          \u001b[0m\u001b[1m \u001b[0m┃\u001b[1m \u001b[0m\u001b[1m      Param #\u001b[0m\u001b[1m \u001b[0m┃\n",
       "┡━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━━━━━━━━━━╇━━━━━━━━━━━━━━━┩\n",
       "│ conv2d_5 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m32\u001b[0m)     │           \u001b[38;5;34m832\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_5 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m32\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout (\u001b[38;5;33mDropout\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m24\u001b[0m, \u001b[38;5;34m32\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_6 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m64\u001b[0m)     │        \u001b[38;5;34m51,264\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_6 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m64\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_1 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m20\u001b[0m, \u001b[38;5;34m64\u001b[0m)     │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_7 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │       \u001b[38;5;34m204,928\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_7 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_2 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m16\u001b[0m, \u001b[38;5;34m128\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ conv2d_8 (\u001b[38;5;33mConv2D\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m256\u001b[0m)    │       \u001b[38;5;34m819,456\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ leaky_re_lu_8 (\u001b[38;5;33mLeakyReLU\u001b[0m)       │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m256\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_3 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m12\u001b[0m, \u001b[38;5;34m256\u001b[0m)    │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ flatten (\u001b[38;5;33mFlatten\u001b[0m)               │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m36864\u001b[0m)          │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dropout_4 (\u001b[38;5;33mDropout\u001b[0m)             │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m36864\u001b[0m)          │             \u001b[38;5;34m0\u001b[0m │\n",
       "├─────────────────────────────────┼────────────────────────┼───────────────┤\n",
       "│ dense_1 (\u001b[38;5;33mDense\u001b[0m)                 │ (\u001b[38;5;45mNone\u001b[0m, \u001b[38;5;34m1\u001b[0m)              │        \u001b[38;5;34m36,865\u001b[0m │\n",
       "└─────────────────────────────────┴────────────────────────┴───────────────┘\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Total params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">1,113,345</span> (4.25 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Total params: \u001b[0m\u001b[38;5;34m1,113,345\u001b[0m (4.25 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">1,113,345</span> (4.25 MB)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Trainable params: \u001b[0m\u001b[38;5;34m1,113,345\u001b[0m (4.25 MB)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/html": [
       "<pre style=\"white-space:pre;overflow-x:auto;line-height:normal;font-family:Menlo,'DejaVu Sans Mono',consolas,'Courier New',monospace\"><span style=\"font-weight: bold\"> Non-trainable params: </span><span style=\"color: #00af00; text-decoration-color: #00af00\">0</span> (0.00 B)\n",
       "</pre>\n"
      ],
      "text/plain": [
       "\u001b[1m Non-trainable params: \u001b[0m\u001b[38;5;34m0\u001b[0m (0.00 B)\n"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "discriminator.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(28, 28, 1)"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "img = img[0]\n",
    "img.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 29ms/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([[0.50325537]], dtype=float32)"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "discriminator.predict(np.expand_dims(img,0))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Construct Training Loop**\n",
    "In this section, we will establish the training loop for our Generative Adversarial Neural Network (GANN), which consists of a Generator and a Discriminator. The training loop includes the following key components:\n",
    "\n",
    "* Loss Functions and Optimizers:\n",
    "    * We will use the Adam optimizer for both models, as it adapts the learning rate based on the average of recent gradients.\n",
    "    * The Binary Cross-Entropy loss will quantify the difference between the predicted and actual labels, providing a measure for our models to optimize against.\n",
    "\n",
    "* Subclassing the Model:\n",
    "    * We will create a custom model class, FashionGAN, that encapsulates both the Generator and Discriminator.\n",
    "    * This class will include methods for compiling the model and executing the training step.\n",
    "\n",
    "* Training Step:\n",
    "    * The Discriminator is trained to differentiate between real and fake images, while the Generator is trained to create images that fool the Discriminator.\n",
    "\n",
    "* Model Monitoring:\n",
    "    * We will implement a callback class, ModelMonitor, to save generated images at the end of each epoch, allowing us to visualize the Generator's progress.\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "# setting up the Losses and Optimizers\n",
    "# Adam is going to be the optimizer for both\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "# Binary cross entropy is going to be the loss for both \n",
    "from tensorflow.keras.losses import BinaryCrossentropy\n",
    "\n",
    "#creating instances of our optimizers and entropy\n",
    "g_opt = Adam(learning_rate=0.0001) \n",
    "d_opt = Adam(learning_rate=0.00001) \n",
    "g_loss = BinaryCrossentropy()\n",
    "d_loss = BinaryCrossentropy()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "#building a subclassed model \n",
    "# Importing the base model class to subclass our training step \n",
    "from tensorflow.keras.models import Model\n",
    "\n",
    "class FashionGAN(Model): \n",
    "    def __init__(self, generator, discriminator, *args, **kwargs):\n",
    "        # Pass through args and kwargs to base class \n",
    "        super().__init__(*args, **kwargs)\n",
    "        \n",
    "        # Create attributes for gen and disc\n",
    "        self.generator = generator \n",
    "        self.discriminator = discriminator \n",
    "        \n",
    "    def compile(self, g_opt, d_opt, g_loss, d_loss, *args, **kwargs): \n",
    "        # Compile with base class\n",
    "        super().compile(*args, **kwargs)\n",
    "        \n",
    "        # Create attributes for losses and optimizers\n",
    "        self.g_opt = g_opt\n",
    "        self.d_opt = d_opt\n",
    "        self.g_loss = g_loss\n",
    "        self.d_loss = d_loss \n",
    "\n",
    "    def train_step(self, batch):\n",
    "        # Get the data \n",
    "        real_images = batch\n",
    "        fake_images = self.generator(tf.random.normal((128, 128, 1)), training=False)\n",
    "        \n",
    "        # Train the discriminator\n",
    "        with tf.GradientTape() as d_tape: \n",
    "            # Pass the real and fake images to the discriminator model\n",
    "            yhat_real = self.discriminator(real_images, training=True) \n",
    "            yhat_fake = self.discriminator(fake_images, training=True)\n",
    "            yhat_realfake = tf.concat([yhat_real, yhat_fake], axis=0)\n",
    "            \n",
    "            # Create labels for real and fakes images\n",
    "            y_realfake = tf.concat([tf.zeros_like(yhat_real), tf.ones_like(yhat_fake)], axis=0)\n",
    "            \n",
    "            # Add some noise to the TRUE outputs\n",
    "            noise_real = 0.15*tf.random.uniform(tf.shape(yhat_real))\n",
    "            noise_fake = -0.15*tf.random.uniform(tf.shape(yhat_fake))\n",
    "            y_realfake += tf.concat([noise_real, noise_fake], axis=0)\n",
    "            \n",
    "            # Calculate loss - BINARYCROSS \n",
    "            total_d_loss = self.d_loss(y_realfake, yhat_realfake)\n",
    "            \n",
    "        # Apply backpropagation - nn learn \n",
    "        dgrad = d_tape.gradient(total_d_loss, self.discriminator.trainable_variables) \n",
    "        self.d_opt.apply_gradients(zip(dgrad, self.discriminator.trainable_variables))\n",
    "        \n",
    "        # Train the generator \n",
    "        with tf.GradientTape() as g_tape: \n",
    "            # Generate some new images\n",
    "            gen_images = self.generator(tf.random.normal((128,128,1)), training=True)\n",
    "                                        \n",
    "            # Create the predicted labels\n",
    "            predicted_labels = self.discriminator(gen_images, training=False)\n",
    "                                        \n",
    "            # Calculate loss - trick to training to fake out the discriminator\n",
    "            total_g_loss = self.g_loss(tf.zeros_like(predicted_labels), predicted_labels) \n",
    "            \n",
    "        # Apply backprop\n",
    "        ggrad = g_tape.gradient(total_g_loss, self.generator.trainable_variables)\n",
    "        self.g_opt.apply_gradients(zip(ggrad, self.generator.trainable_variables))\n",
    "        \n",
    "        return {\"d_loss\":total_d_loss, \"g_loss\":total_g_loss}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Create instance of subclassed model\n",
    "fashongnn = FashionGAN(generator, discriminator)\n",
    "\n",
    "# Compile the model\n",
    "fashongnn.compile(g_opt, d_opt, g_loss, d_loss)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "#optional step, Note the \"directory path\" need to be modified.\n",
    "#building our callback\n",
    "import os\n",
    "from tensorflow.keras.preprocessing.image import array_to_img\n",
    "from tensorflow.keras.callbacks import Callback\n",
    "\n",
    "class ModelMonitor(Callback):\n",
    "    def __init__(self, num_img=3, latent_dim=128):\n",
    "        self.num_img = num_img\n",
    "        self.latent_dim = latent_dim\n",
    "\n",
    "    def on_epoch_end(self, epoch, logs=None):\n",
    "        random_latent_vectors = tf.random.uniform((self.num_img, self.latent_dim,1))\n",
    "        generated_images = self.model.generator(random_latent_vectors)\n",
    "        generated_images *= 255\n",
    "        generated_images.numpy()\n",
    "        for i in range(self.num_img):\n",
    "            img = array_to_img(generated_images[i])\n",
    "            img.save(os.path.join('images', f'generated_img_{epoch}_{i}.png'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/5\n",
      "\u001b[1m469/469\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m370s\u001b[0m 789ms/step - d_loss: 0.5355 - g_loss: 1.1180\n",
      "Epoch 2/5\n",
      "\u001b[1m469/469\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m370s\u001b[0m 789ms/step - d_loss: 0.3408 - g_loss: 3.9687\n",
      "Epoch 3/5\n",
      "\u001b[1m469/469\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m367s\u001b[0m 782ms/step - d_loss: 0.2844 - g_loss: 5.8856\n",
      "Epoch 4/5\n",
      "\u001b[1m469/469\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m366s\u001b[0m 780ms/step - d_loss: 0.2789 - g_loss: 5.8143\n",
      "Epoch 5/5\n",
      "\u001b[1m469/469\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m366s\u001b[0m 780ms/step - d_loss: 0.4423 - g_loss: 2.3507\n"
     ]
    }
   ],
   "source": [
    "#training the model:\n",
    "#due to resource constraints we will be training it for 5 epochs\n",
    "#recommended epochs should be around 2000!\n",
    "hist = fashongnn.fit(ds, epochs=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#checking the performance\n",
    "\n",
    "#for loss\n",
    "plt.suptitle('Loss')\n",
    "plt.plot(hist.history['d_loss'], label='d_loss')\n",
    "plt.plot(hist.history['g_loss'], label='g_loss')\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "**Test Out the Generator**\n",
    "* Load the pre-trained weights for the generator model.\n",
    "* Generate a set of images using random input noise.\n",
    "* Visualize the generated images in a grid format.\n",
    "* Save the generator and discriminator models for future use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\u001b[1m1/1\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 38ms/step\n"
     ]
    }
   ],
   "source": [
    "generator = tf.keras.models.load_model('generatormodel.h5')\n",
    "imgs = generator.predict(tf.random.normal((16,128,1)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x1000 with 16 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax = plt.subplots(ncols=4, nrows=4, figsize=(10,10))\n",
    "for r in range(4): \n",
    "    for c in range(4): \n",
    "        ax[r][c].imshow(imgs[(r+1)*(c+1)-1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#saving our model!\n",
    "generator.save('generator.h5')\n",
    "discriminator.save('discriminator.h5')"
   ]
  }
 ],
 "metadata": {
  "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": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}