{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# unit 1.2 - Binary net in PyTorch with manual weights\n",
    "\n",
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://githubtocolab.com/culurciello/deep-learning-course-source/blob/main/source/lectures/12-binary-net-pytorch-manual.ipynb)\n",
    "\n",
    "Now we will learn what is a \"proper\" way to define this super simple AND neural network in PyTorch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d7fbcff1",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "from torch import nn\n",
    "import torch.nn.functional as F"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "44eee2e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# professional way to define model in pytorch:\n",
    "\n",
    "class Logic_net(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(Logic_net, self).__init__()\n",
    "        self.l1 = nn.Linear(2, 2)\n",
    "        self.l2 = nn.Linear(2, 1)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = F.relu(self.l1(x))\n",
    "        x = self.l2(x) \n",
    "        return x\n",
    "    \n",
    "model = Logic_net()\n",
    "\n",
    "# AND model weights:\n",
    "model.l1.weight = nn.Parameter(torch.Tensor([[0,0],[0.45,0.45]]))\n",
    "model.l1.bias = nn.Parameter(torch.Tensor([0,0]))\n",
    "model.l2.weight = nn.Parameter(torch.Tensor([[0,1]]))\n",
    "model.l2.bias = nn.Parameter(torch.Tensor([0]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "04f83446",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "out: 0.0 0.44999998807907104 0.44999998807907104 0.8999999761581421\n"
     ]
    }
   ],
   "source": [
    "# AND model test:\n",
    "i1 = torch.Tensor([0,0])\n",
    "o1 = model(i1)\n",
    "i2 = torch.Tensor([0,1])\n",
    "o2 = model(i2)\n",
    "i3 = torch.Tensor([1,0])\n",
    "o3 = model(i3)\n",
    "i4 = torch.Tensor([1,1])\n",
    "o4 = model(i4)\n",
    "\n",
    "print(\"out:\", o1.item(),o2.item(),o3.item(),o4.item())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "623db121",
   "metadata": {},
   "outputs": [],
   "source": [
    "# another easier way to define model:\n",
    "\n",
    "model = torch.nn.Sequential(\n",
    "    torch.nn.Linear(2, 2),\n",
    "    torch.nn.ReLU(),\n",
    "    torch.nn.Linear(2, 1),\n",
    ")\n",
    "\n",
    "# print(model)\n",
    "# print(model[0].weight)\n",
    "\n",
    "# a AND model weights:\n",
    "model[0].weight = nn.Parameter(torch.Tensor([[0.3,-0.2],[1,1]]))\n",
    "model[0].bias = nn.Parameter(torch.Tensor([-0.5,-0.5]))\n",
    "model[2].weight = nn.Parameter(torch.Tensor([[-0.2,0.9]]))\n",
    "model[2].bias = nn.Parameter(torch.Tensor([-0.2]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d8f2fee9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "out: -0.20000000298023224 0.2499999850988388 0.2499999850988388 1.1499998569488525\n"
     ]
    }
   ],
   "source": [
    "# AND model test:\n",
    "i1 = torch.Tensor([0,0])\n",
    "o1 = model(i1)\n",
    "i2 = torch.Tensor([0,1])\n",
    "o2 = model(i2)\n",
    "i3 = torch.Tensor([1,0])\n",
    "o3 = model(i3)\n",
    "i4 = torch.Tensor([1,1])\n",
    "o4 = model(i4)\n",
    "\n",
    "print(\"out:\", o1.item(),o2.item(),o3.item(),o4.item())"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c4601502",
   "metadata": {},
   "source": [
    "## HOMEWORK\n",
    "\n",
    "Try to make the XOR or XNOR neural network on your own.\n",
    "\n",
    "Tip: think about decomposing the XOR function: XOR(x1,x2) = $x1*x2 + NOT(x1)*NOT(+x2)$\n",
    "\n",
    "Note that $x1*x2$ = AND(x1, x2)\n",
    "\n",
    "Can it be solved with 1 neuron?"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}