Sine Function Approximation in PyTorch.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Sine Function Approximation in PyTorch.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true,
      "authorship_tag": "ABX9TyMw1E2kKQXtqt3H4ylgtE0P",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/smlra-kjsce/PyTorch-101/blob/main/Sine%20Function%20Approximation%20in%20PyTorch.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dR0RKxShTEiT"
      },
      "source": [
        "## Import Libraries"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pNssNOferi7B"
      },
      "source": [
        "import torch\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "from torch import nn, optim\n",
        "from torch.nn import functional as F\n",
        "from torch.utils.data import TensorDataset, DataLoader\n",
        "\n",
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "device = torch.device(\"cuda\") if torch.cuda.is_available() else torch.device(\"cpu\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "MAxg-qX-TIEO"
      },
      "source": [
        "## Define some hyperparameters"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Dw_J3We1XOl5"
      },
      "source": [
        "LR = 1e-6\n",
        "MAX_EPOCH = 10\n",
        "BATCH_SIZE = 512"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "1Hrjs1vgTKf2"
      },
      "source": [
        "## Defining our model to approximate sine curve"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uOf7aGyEXSc5"
      },
      "source": [
        "class SineApproximator(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(SineApproximator, self).__init__()\n",
        "        self.regressor = nn.Sequential(nn.Linear(1, 1024),\n",
        "                                       nn.ReLU(inplace=True),\n",
        "                                       nn.Linear(1024, 1024),\n",
        "                                       nn.ReLU(inplace=True),\n",
        "                                       nn.Linear(1024, 1))\n",
        "    def forward(self, x):\n",
        "        output = self.regressor(x)\n",
        "        return output\n",
        "\n",
        "    ## Since I have inherited nn.Module, backward() function is already implemented and I don't need to define it again."
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "KXgTdqUDXWu6"
      },
      "source": [
        "X = np.random.rand(10**6) * 2 * np.pi\n",
        "y = np.sin(X)\n",
        "\n",
        "X_train, X_val, y_train, y_val = map(torch.tensor, train_test_split(X, y, test_size=0.2))\n",
        "train_dataloader = DataLoader(TensorDataset(X_train.unsqueeze(1), y_train.unsqueeze(1)), batch_size=BATCH_SIZE,\n",
        "                              pin_memory=True, shuffle=True)\n",
        "val_dataloader = DataLoader(TensorDataset(X_val.unsqueeze(1), y_val.unsqueeze(1)), batch_size=BATCH_SIZE,\n",
        "                            pin_memory=True, shuffle=True)\n",
        "\n",
        "model = SineApproximator().to(device)\n",
        "optimizer = optim.Adam(model.parameters(), lr=LR)\n",
        "criterion = nn.MSELoss(reduction=\"mean\")"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "_BjPn291YbNL",
        "outputId": "2c65a674-4a37-4408-9b51-e7f12fe66c66"
      },
      "source": [
        "plt.scatter(X,y)\n",
        "plt.title('Total Dataset for Sine Approximation');"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "M9Yil3aeYpdK",
        "outputId": "749eda5d-479b-42ef-9791-9fdba314500f"
      },
      "source": [
        "plt.scatter(X_train.cpu(),y_train.cpu())\n",
        "plt.title('train Dataset for Sine Approximation');"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        },
        "id": "dV15m7-jYrEh",
        "outputId": "4d974912-ff68-4dd0-cb9e-cdab6967be62"
      },
      "source": [
        "plt.scatter(X_val.cpu(),y_val.cpu())\n",
        "plt.title('val Dataset for Sine Approximation');"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Pbhp7wxOTQH8"
      },
      "source": [
        "## Training our model"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "vAwzbZ3-Xkid",
        "outputId": "8c322c26-8a54-4fcd-e3e3-2c6b7eb08449"
      },
      "source": [
        "train_loss_list = list()\n",
        "val_loss_list = list()\n",
        "for epoch in range(MAX_EPOCH):\n",
        "    print(\"epoch %d / %d\" % (epoch+1, MAX_EPOCH))\n",
        "    model.train()\n",
        "    # training loop\n",
        "    temp_loss_list = list()\n",
        "    for X_train, y_train in train_dataloader:\n",
        "        X_train = X_train.type(torch.float32).to(device)\n",
        "        y_train = y_train.type(torch.float32).to(device)\n",
        "\n",
        "        optimizer.zero_grad()\n",
        "\n",
        "        score = model(X_train)\n",
        "        loss = criterion(input=score, target=y_train)\n",
        "        loss.backward()\n",
        "\n",
        "        optimizer.step()\n",
        "\n",
        "        temp_loss_list.append(loss.detach().cpu().numpy())\n",
        "    \n",
        "    temp_loss_list = list()\n",
        "    for X_train, y_train in train_dataloader:\n",
        "        X_train = X_train.type(torch.float32).to(device)\n",
        "        y_train = y_train.type(torch.float32).to(device)\n",
        "\n",
        "        score = model(X_train)\n",
        "        loss = criterion(input=score, target=y_train)\n",
        "\n",
        "        temp_loss_list.append(loss.detach().cpu().numpy())\n",
        "    \n",
        "    train_loss_list.append(np.average(temp_loss_list))\n",
        "\n",
        "    # validation\n",
        "    model.eval()\n",
        "    \n",
        "    temp_loss_list = list()\n",
        "    for X_val, y_val in val_dataloader:\n",
        "        X_val = X_val.type(torch.float32).to(device)\n",
        "        y_val = y_val.type(torch.float32).to(device)\n",
        "\n",
        "        score = model(X_val)\n",
        "        loss = criterion(input=score, target=y_val)\n",
        "\n",
        "        temp_loss_list.append(loss.detach().cpu().numpy())\n",
        "    \n",
        "    val_loss_list.append(np.average(temp_loss_list))\n",
        "\n",
        "    print(\"\\ttrain loss: %.5f\" % train_loss_list[-1])\n",
        "    print(\"\\tval loss: %.5f\" % val_loss_list[-1])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "epoch 1 / 10\n",
            "\ttrain loss: 0.09842\n",
            "\tval loss: 0.09853\n",
            "epoch 2 / 10\n",
            "\ttrain loss: 0.07372\n",
            "\tval loss: 0.07380\n",
            "epoch 3 / 10\n",
            "\ttrain loss: 0.05159\n",
            "\tval loss: 0.05164\n",
            "epoch 4 / 10\n",
            "\ttrain loss: 0.03287\n",
            "\tval loss: 0.03290\n",
            "epoch 5 / 10\n",
            "\ttrain loss: 0.01913\n",
            "\tval loss: 0.01915\n",
            "epoch 6 / 10\n",
            "\ttrain loss: 0.01024\n",
            "\tval loss: 0.01024\n",
            "epoch 7 / 10\n",
            "\ttrain loss: 0.00516\n",
            "\tval loss: 0.00516\n",
            "epoch 8 / 10\n",
            "\ttrain loss: 0.00249\n",
            "\tval loss: 0.00250\n",
            "epoch 9 / 10\n",
            "\ttrain loss: 0.00116\n",
            "\tval loss: 0.00116\n",
            "epoch 10 / 10\n",
            "\ttrain loss: 0.00053\n",
            "\tval loss: 0.00053\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Y60Ym2BsTWhr"
      },
      "source": [
        "## Plotting Loss"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "G--C0PBRsNMx",
        "outputId": "f7c773e9-9f2d-4362-9e18-97c6227e8042"
      },
      "source": [
        "train_plot = plt.plot(train_loss_list, label=\"Training Loss\")\n",
        "val_plot = plt.plot(val_loss_list, label=\"Validation Loss\")\n",
        "plt.legend(loc=\"upper right\");"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAD4CAYAAADiry33AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deVxVdf7H8deHy76IoLiBBm6YioLinonZYptm6qRpaZapkzXZNGVN2zTjVPOzzSlrWixLy8zSsdXKNC3HBRUXRBMRFVdARWW/8P39wdWUUFEvnMvl83w8fHTvOd9zed9rvu/he889R4wxKKWUcl8eVgdQSilVtbTolVLKzWnRK6WUm9OiV0opN6dFr5RSbs7T6gDl1a9f30RGRlodQymlapS1a9dmGWPCKlrnckUfGRlJYmKi1TGUUqpGEZFdZ1unUzdKKeXmtOiVUsrNVaroRaS/iGwTkVQRmVzB+itFZJ2I2EVkSLl1o0Rku+PPKGcFV0opVTnnnaMXERvwOnANkAGsEZGFxpgtpw3bDYwGHi63bSjwNBAPGGCtY9sjzomvlLoUxcXFZGRkUFBQYHUUVUm+vr5ERETg5eVV6W0q82FsVyDVGJMGICJzgIHAqaI3xqQ71pWW2/Y64HtjzGHH+u+B/sDHlU6olKoyGRkZBAUFERkZiYhYHUedhzGG7OxsMjIyiIqKqvR2lZm6CQf2nHY/w7GsMiq1rYjcKyKJIpKYmZlZyYdWSl2qgoIC6tWrpyVfQ4gI9erVu+DfwFziw1hjzFvGmHhjTHxYWIWHgSqlqoiWfM1yMX9flSn6vUDT0+5HOJZVxqVse0FM3hHSPpyIPVen/5VS6nSVKfo1QCsRiRIRb2AYsLCSj78IuFZEQkQkBLjWsczpkjauo1nqbLZ/eH9VPLxSysmys7OJjY0lNjaWRo0aER4efup+UVHRObdNTEzkgQceOO/P6Nmzp1OyLl26lJtuuskpj2WF834Ya4yxi8hEygraBswwxiSLyLNAojFmoYh0AeYDIcDNIvI3Y0w7Y8xhEfk7ZW8WAM+e/GDW2WK7XcW3q4Zz/YHZZKz8jIjug6vixyilnKRevXokJSUB8MwzzxAYGMjDD/924J7dbsfTs+KKio+PJz4+/rw/Y8WKFc4JW8NVao7eGPO1Maa1MaaFMWaKY9lTxpiFjttrjDERxpgAY0w9Y0y707adYYxp6fjzXtU8jbJ5q66jX+BXLiNg0UMUHcuqqh+llKoio0ePZvz48XTr1o1HHnmE1atX06NHD+Li4ujZsyfbtm0DztzDfuaZZxgzZgwJCQk0b96cadOmnXq8wMDAU+MTEhIYMmQIbdq0YcSIEZy8ut7XX39NmzZt6Ny5Mw888MAF7bl//PHHxMTE0L59ex599FEASkpKGD16NO3btycmJoaXX34ZgGnTptG2bVs6dOjAsGHDLv3FugAud66bS1EvOIht10wj8rtbSftgPG0mzrM6klI1xt++SGbLvmNOfcy2Terw9M3tzj/wNBkZGaxYsQKbzcaxY8dYvnw5np6e/PDDDzz++ON89tlnv9tm69atLFmyhOPHjxMdHc2ECRN+d5z5+vXrSU5OpkmTJvTq1YtffvmF+Ph4xo0bx7Jly4iKimL48OGVzrlv3z4effRR1q5dS0hICNdeey0LFiygadOm7N27l82bNwNw9OhRAJ5//nl27tyJj4/PqWXVxSWOunGmnr0S+KHBXbTJ+p5dy2ZZHUcpdYGGDh2KzWYDICcnh6FDh9K+fXsmTZpEcnJyhdvceOON+Pj4UL9+fRo0aMDBgwd/N6Zr165ERETg4eFBbGws6enpbN26lebNm586Jv1Cin7NmjUkJCQQFhaGp6cnI0aMYNmyZTRv3py0tDTuv/9+vv32W+rUqQNAhw4dGDFiBLNmzTrrlFRVcas9+pN6jfoHyS8uJWLJZApi+uEb0tjqSEq5vAvd864qAQEBp24/+eST9O3bl/nz55Oenk5CQkKF2/j4+Jy6bbPZsNvtFzXGGUJCQtiwYQOLFi3izTffZO7cucyYMYOvvvqKZcuW8cUXXzBlyhQ2bdpUbYXvdnv0AMGBfuTd8Bq+pQXsnjkWHHNxSqmaJScnh/Dwsu9Yvv/++05//OjoaNLS0khPTwfgk08+qfS2Xbt25aeffiIrK4uSkhI+/vhj+vTpQ1ZWFqWlpQwePJh//OMfrFu3jtLSUvbs2UPfvn154YUXyMnJ4cSJE05/PmfjlkUP0KVLD35oMo7WR5ezc/E7VsdRSl2ERx55hMcee4y4uLgq2QP38/Nj+vTp9O/fn86dOxMUFERwcHCFYxcvXkxERMSpP+np6Tz//PP07duXjh070rlzZwYOHMjevXtJSEggNjaWkSNH8txzz1FSUsLIkSOJiYkhLi6OBx54gLp16zr9+ZyNGBfb242PjzfOuvDIifxCdvxfH5qX7sJz4kr86l/mlMdVyl2kpKRw+eWXWx3DUidOnCAwMBBjDPfddx+tWrVi0qRJVsc6p4r+3kRkrTGmwmNO3XaPHiDQz4fSgdOxmRL2zrxHp3CUUr/z9ttvExsbS7t27cjJyWHcuHFWR3I6ty56gLiOnVja7H5aHl9N6jf/tjqOUsrFTJo0iaSkJLZs2cLs2bPx9/e3OpLTuX3RA1x1x2QSbR1psnoKxw9stzqOUkpVq1pR9L7eXvgOnk6JETI/uBtKy582Xyml3FetKHqA9m3bs7zFn2met4FfF/6f1XGUUqra1JqiB+g3fBIrPbvQLGkqOXu2nH8DpZRyA7Wq6H28PAkZNp0C48Xh2WOgpGq+GaeUqpy+ffuyaNGZZy5/5ZVXmDBhwlm3SUhI4OQh2DfccEOF54155plnmDp16jl/9oIFC9iy5bcdvqeeeooffvjhQuJXyBVPaVyrih4gumVrVrZ5nKiCFFI+n2J1HKVqteHDhzNnzpwzls2ZM6fS55z5+uuvL/qLR+WL/tlnn+Xqq6++qMdydbWu6AGuHjqBX7x70TL5VQ6nrbc6jlK11pAhQ/jqq69OXWgkPT2dffv20bt3byZMmEB8fDzt2rXj6aefrnD7yMhIsrLKTkk+ZcoUWrduzRVXXHHqdMZQdpx8ly5d6NixI4MHDyYvL48VK1awcOFC/vKXvxAbG8uOHTsYPXo08+aVnfF28eLFxMXFERMTw5gxYygsLDz1855++mk6depETEwMW7durfRztfKUxm55UrPz8fS00XjEdHJmXEnunHsIeWQF4ulz/g2VcmffTIYDm5z7mI1i4Prnz7o6NDSUrl278s033zBw4EDmzJnDH/7wB0SEKVOmEBoaSklJCf369WPjxo106NChwsdZu3Ytc+bMISkpCbvdTqdOnejcuTMAt956K2PHjgXgiSee4N133+X+++9nwIAB3HTTTQwZMuSMxyooKGD06NEsXryY1q1bc+edd/LGG2/w4IMPAlC/fn3WrVvH9OnTmTp1Ku+8c/5TrFh9SuNauUcP0PyySNZ1eJrLilJJmVvx3oJSquqdPn1z+rTN3Llz6dSpE3FxcSQnJ58xzVLe8uXLGTRoEP7+/tSpU4cBAwacWrd582Z69+5NTEwMs2fPPuupjk/atm0bUVFRtG7dGoBRo0axbNmyU+tvvfVWADp37nzqZGjnY/UpjWvlHv1J/QaN4aftX9Hr1/+QuW0gYdE9rI6klHXOseddlQYOHMikSZNYt24deXl5dO7cmZ07dzJ16lTWrFlDSEgIo0ePpqCg4KIef/To0SxYsICOHTvy/vvvs3Tp0kvKe/J0x8441XF1ndK41u7RA9g8hOZ3/JssU5fCT8dhivOtjqRUrRMYGEjfvn0ZM2bMqb35Y8eOERAQQHBwMAcPHuSbb74552NceeWVLFiwgPz8fI4fP84XX3xxat3x48dp3LgxxcXFzJ49+9TyoKAgjh8//rvHio6OJj09ndTUVAA+/PBD+vTpc0nP0epTGtfqPXqApuHhLI7/B/3W/pHkjybTbtSrVkdSqtYZPnw4gwYNOjWF07FjR+Li4mjTpg1NmzalV69e59y+U6dO3HbbbXTs2JEGDRrQpUuXU+v+/ve/061bN8LCwujWrdupch82bBhjx45l2rRppz6EBfD19eW9995j6NCh2O12unTpwvjx4y/o+Zw8pfFJn3766alTGhtjuPHGGxk4cCAbNmzgrrvuotTxbf3TT2mck5ODMcYppzR269MUV5YxhiVTbyfhxDccGvpfGrW/tHdvpWoKPU1xzaSnKb4IIkKbUdPYR33M/PGUFlTflV+UUqqqadE7NGkQRmqPF2hcso+UWQ9ZHUcppZxGi/40fa67lR/qDKJdxifsXfet1XGUqhauNn2rzu1i/r606E8jInQY/RK7aIzXlxOx5136FxWUcmW+vr5kZ2dr2dcQxhiys7Px9fW9oO1q/VE35TUIDeXnPi/SY+kItn7wJ9qNn2l1JKWqTEREBBkZGWRmZlodRVWSr6/vGUf0VIYWfQWu6Hsj320cxrUHPmb3ygU0636L1ZGUqhJeXl5ERUVZHUNVMZ26OYv40f/HDpoSsGgSRcezrY6jlFIXTYv+LEKDg8i8Zhp1SnNInflHq+MopdRF06I/h+69ruLHBnfSNutb0pZ9bHUcpZS6KFr059Fj9D/ZKs0JWfIoBUcPWB1HKaUuWKWKXkT6i8g2EUkVkckVrPcRkU8c61eJSKRjuZeIzBSRTSKSIiKPOTd+1asT4E/uDa/hX5rLzvfHgR6GppSqYc5b9CJiA14HrgfaAsNFpG25YXcDR4wxLYGXgRccy4cCPsaYGKAzMO7km0BN0rlLL5Y0GcvlR5eSuniG1XGUUuqCVGaPviuQaoxJM8YUAXOAgeXGDAROHnA+D+gnIgIYIEBEPAE/oAg45pTk1az3nX9js0c0DX9+krysPVbHUUqpSqtM0YcDpzdbhmNZhWOMMXYgB6hHWennAvuB3cBUY8zhS8xsiQA/H0oHTMfTFLNn5j06haOUqjGq+sPYrkAJ0ASIAv4sIs3LDxKRe0UkUUQSXfkbeh1i41l22X1EH1/Jtm+nWx1HKaUqpTJFvxdoetr9CMeyCsc4pmmCgWzgduBbY0yxMeYQ8Avwu/MlG2PeMsbEG2Piw8LCLvxZVKM+I/9Kkq09Eaue5diBHVbHUUqp86pM0a8BWolIlIh4A8OAheXGLARGOW4PAX40ZWdJ2g1cBSAiAUB3YKszglvF19sL78FvYgwc/OBucFwZRimlXNV5i94x5z4RWASkAHONMcki8qyInLzU+rtAPRFJBR4CTh6C+ToQKCLJlL1hvGeM2ejsJ1Hd2raNYUXLh2iVt56UL16yOo5SSp2TXkrwIhUVl7DhhWuIsW+i4J5l1I3Qy7EppayjlxKsAt5eNuoOe4NC40X2rLsxJXarIymlVIW06C9Bq5bRrL58Mi0Kktny+XNWx1FKqQpp0V+ivkPuY6V3D1olv0J2WpLVcZRS6ne06C+Rp6eNhre/wQnjx/E5YzH2IqsjKaXUGbTonSAqMor1HZ4isuhXkuc+Y3UcpZQ6gxa9k/QdNJaffROI3vYmB7etsjqOUkqdokXvJB4eQtSd0zlKEEWf3ktpUb7VkZRSCtCid6rwJuEkx0+hqT2dLe/fZ3UcpZQCtOidrs9NI/g+5Dba7/uM7YvfszqOUkpp0TubiND9nlfY7NGG8OWTOZy+2epISqlaTou+CgQF+OM97AMKjRe5s0ZQUphrdSSlVC2mRV9FWreOZmPXfxFevIuUGeOtjqOUqsW06KvQlTcMZ3HYSNofXMiv375pdRylVC2lRV+FRIRe90wlyRZDs5VPkpW23upISqlaSIu+ivn7+hI88n2OGz8KZt+BPb9GXhtdKVWDadFXg6iolmzt9RKN7Rlse0cvLK6Uql5a9NWk97VD+LHR3bTLXsTWr6ZZHUcpVYto0Vej3nc/z1rPOKIS/87BX1dbHUcpVUto0VcjX28vwkZ9wBETRMknoyg6ccTqSEqpWkCLvpo1a9qMnQn/poH9AKnv3KXz9UqpKqdFb4EefW9iSfh42h5dwpb/TrU6jlLKzWnRW6TPXX9ntVdXWq1/jv1bfrE6jlLKjWnRW8Tby5Pwu94nU0KQeXdRcCzb6khKKTelRW+h8Cbh7On3BqElWex8506dr1dKVQkteot1630tP132AJcf+5nN86ZYHUcp5Ya06F1Awp1PstKnF9GbXyJjwxKr4yil3IwWvQvw8rQROWYGB6Q+PgvuIf/oIasjKaXciBa9i2jUsBGZ/f9DndKj7HpnJJSWWh1JKeUmtOhdSKfufVne4iHanFjFxrnPWB1HKeUmtOhdTN8Rj7HCL4F2KdPYtW6R1XGUUm5Ai97F2GwetLrnXTKkMYFfjCP38D6rIymlajgtehcUVq8+R296m4DSE2S8MwJTYrc6klKqBqtU0YtIfxHZJiKpIjK5gvU+IvKJY/0qEYk8bV0HEfmfiCSLyCYR8XVefPfVMf4KVkQ/RnTeOjZ99Fer4yilarDzFr2I2IDXgeuBtsBwEWlbbtjdwBFjTEvgZeAFx7aewCxgvDGmHZAAFDstvZtLuG0SywOuoX3qf9i56gur4yilaqjK7NF3BVKNMWnGmCJgDjCw3JiBwEzH7XlAPxER4FpgozFmA4AxJtsYU+Kc6O7Pw+ZBu3veZqdHBHW/uY/jmbutjqSUqoEqU/ThwJ7T7mc4llU4xhhjB3KAekBrwIjIIhFZJyKPVPQDROReEUkUkcTMzMwLfQ5uLTQkhIJbZuBtCjjw7u2YEv2FSCl1Yar6w1hP4ApghOO/g0SkX/lBxpi3jDHxxpj4sLCwKo5U87Tr2JU17Z6kVcEmNn5Y4XulUkqdVWWKfi/Q9LT7EY5lFY5xzMsHA9mU7f0vM8ZkGWPygK+BTpcaujbqM3QiPwXdSMf0GaT+8rnVcZRSNUhlin4N0EpEokTEGxgGLCw3ZiEwynF7CPCjMcYAi4AYEfF3vAH0AbY4J3rtIiLE3vMm2yWSsO/v5+j+NKsjKaVqiPMWvWPOfSJlpZ0CzDXGJIvIsyIywDHsXaCeiKQCDwGTHdseAV6i7M0iCVhnjPnK+U+jdggOrkPJkPexmRKy3rud0uJCqyMppWoAMS52sYv4+HiTmJhodQyX9tP8t+iz4S8kRYwk9p7XrY6jlHIBIrLWGBNf0Tr9ZmwNdOUtY/mp7iBiM2bx69KPrY6jlHJxWvQ1kIjQeezrbPVoSeOlD3E4Y5vVkZRSLkyLvoYKDAjA87aZlBrh6MwRlBQVWB1JKeWitOhrsJbR7dkU/xzNi7ez6b2JVsdRSrkoLfoartdNd7Ik9DZi93/K1h9mnn8DpVSto0Vfw4kI3ca+SrKtDRE/P0rWrmSrIymlXIwWvRvw9/Mj4PYPKDaenPhwBPaCXKsjKaVciBa9m4hsEU1Kj/8j0r6T5BkTrI6jlHIhWvRupGf/4SxpcAcdD/2XLd+8ZXUcpZSL0KJ3Mz3ufpGNnu2JWvUEB3ckWR1HKeUCtOjdjK+PDyF3fEie8aXoo5Hk52RbHUkpZTEtejfU9LLmpPV9jQb2/eyZPhB7YZ7VkZRSFtKid1NdEgawKvaftCzYzNbXh2FK7FZHUkpZRIvejV05aBxLoh6k/bGf2PjOeHCxM5UqpaqHFr2bu2rU0yypN5yO+z9lw8dPWx1HKWUBLXo3JyJcMeE1VgT0o+Ovr7L56zesjqSUqmZa9LWAl6cnsRNnsd4rjjarHme7XnNWqVpFi76W8Pfzp9mEz0izRRLx/Xj2bPrZ6khKqWqiRV+L1Auth//o+WQTTOBnw8nalWJ1JKVUNdCir2UimkWSN3QuxhiKZt7C8ey9VkdSSlUxLfpaqHW7OHb3f4+6JUfIfHMghXk5VkdSSlUhLfpaKrbHNWzo/grNinaw47XBlBYXWR1JKVVFtOhrsR7X387Plz9B27w1bHrjTkxpqdWRlFJVQIu+lutz20MsaTKWjoe/IWnmQ1bHUUpVAS36Wk5E6HP3v/g5+Gbidr3Hhs9esDqSUsrJtOgVHjYPutw3gzW+PYjZ+Bwpiz+wOpJSyom06BUAPt7eRN83lxTPNjRf9hBpid9ZHUkp5SRa9OqUOkF1CLt3Pvs8GhL25Sj2b1trdSSllBNo0aszNGjYGBn5Ofn44DlnKIf3pVkdSSl1ibTo1e9Etogmc+DH+JbmceLdgeQezbQ6klLqEmjRqwq1i+vBtr5v0dC+j4w3bqG4INfqSEqpi1SpoheR/iKyTURSRWRyBet9ROQTx/pVIhJZbn0zETkhIg87J7aqDvEJA1jT6TlaFSSz9fXb9HKEStVQ5y16EbEBrwPXA22B4SLSttywu4EjxpiWwMtA+YOxXwK+ufS4qrpdMfBeljWfRMzx5SS9da9ejlCpGqgye/RdgVRjTJoxpgiYAwwsN2YgMNNxex7QT0QEQERuAXYCyc6JrKpbnzuf4qew4cQd/Iz1Hz1pdRyl1AWqTNGHA3tOu5/hWFbhGGOMHcgB6olIIPAo8Ldz/QARuVdEEkUkMTNTP/hzNSJCr3GvsTKgH3Hb/82mL1+3OpJS6gJU9YexzwAvG2NOnGuQMeYtY0y8MSY+LCysiiOpi+Hp6UnHibNJ8orj8jVPsG25Xo5QqZqiMkW/F2h62v0Ix7IKx4iIJxAMZAPdgH+JSDrwIPC4iEy8xMzKIn5+flz2x8/ZaYuk2eLx7Nq43OpISqlKqEzRrwFaiUiUiHgDw4CF5cYsBEY5bg8BfjRlehtjIo0xkcArwD+NMa85KbuyQEhIKAFj5nOYYII+v52DO/WjF6Vc3XmL3jHnPhFYBKQAc40xySLyrIgMcAx7l7I5+VTgIeB3h2Aq99EkIpL8P3wKxlD6wSCOZerlCJVyZWJc7HC5+Ph4k5iYaHUMVQkbV/5Ay29u54BXU5o8uBjfwLpWR1Kq1hKRtcaY+IrW6Tdj1UXr0P1qNvV8hWbFaex4fTAlxYVWR1JKVUCLXl2Sbtfdzv/aPkW7/EQ2TdfLESrlirTo1SXrfdskloXfS+yRb1n33oNWx1FKlaNFr5ziijEv8Evdm+m8ZybrPn3e6jhKqdNo0Sun8LB50OWP77HWrwexm59n83czz7+RUqpaaNErp/H29iL6vk/Z6tWG1r88xPbV31odSSmFFr1yssDAIMflCBvR6Ou72LN1jdWRlKr1tOiV04U1aIzHnZ+Tjy8+n9xG1t4dVkdSqlbToldVollUNIcHfYxfaR5F71xPhl5oXCnLaNGrKtOmY3fSb/wIL1NEyMc3kPzjR1ZHUqpW0qJXVSqm61UU3bWYDFtT2i2bwLoPH9cvVSlVzbToVZULv6wF4ZOWsDKwH512vM7GV26lMP+41bGUqjW06FW1CAwMouukeSy7bCIxOUvZ+2IfsvemWR1LqVpBi15VGw+bB1feNYXEntNpULwPeTuBHesWWx1LKbenRa+qXdfrbmf/0C/JFX+a/ncoSf/9t9WRlHJrWvTKEq3ax+M3YSkpPh2IXf8Eif8ZR6m92OpYSrklLXplmfoNGtHm4UX8XG8o8fvnkPLidZw4mmV1LKXcjha9spSPtw+9Jr7NL22fplVeEjnTrmDf9iSrYynlVrToleVEhF5/eIiUaz/CtzSPoNn92bLsU6tjKeU2tOiVy+jYqz95o77noEcj2iwey9qPntEvVynlBFr0yqU0jYqm4YNLWRvQm86/vkzStGEUFeRZHUupGk2LXrmcoDp16fTn/7I8YhxxRxex68UEDh/YZXUspWosLXrlkmw2D3rf8y9WdZ1Gk6J0St/sQ1rST1bHUqpG0qJXLq3bDaPYO3ghRXgRPn8wG776j9WRlKpxtOiVy2vdoTteE5ay3bsNHdc8QuLbEym1262OpVSNoUWvaoSwhuG0/PMPrAi5hfi9H5L80g3k5hy2OpZSNYIWvaoxfH196fHA+/wS/RhtchPJfrU3+9M2Wx1LKZenRa9qFBGh1/DJJF89k6DSHAI+uJaUnxdYHUspl6ZFr2qk2N43c/yO78jyqE/r70eT+MkUMMbqWEq5JC16VWM1a9GW+n9aSpJ/T+JT/sXaf4+kuDDf6lhKuRwtelWj1QkOJfbhL/i5yRg6H/6StKlXceRQhtWxlHIplSp6EekvIttEJFVEJlew3kdEPnGsXyUikY7l14jIWhHZ5PjvVc6NrxTYbDauuPdlVsW/RLOiVIreuJKdm36xOpZSLuO8RS8iNuB14HqgLTBcRNqWG3Y3cMQY0xJ4GXjBsTwLuNkYEwOMAj50VnClyut2093svmU+xgiN5t1C0rczrI6klEuozB59VyDVGJNmjCkC5gADy40ZCMx03J4H9BMRMcasN8bscyxPBvxExMcZwZWqSHTcFXjcu4SdXi2JXTmJ1e8+hCktsTqWUpaqTNGHA3tOu5/hWFbhGGOMHcgB6pUbMxhYZ4wpvLioSlVOgybNaP7nxawMvpGue95l44s3k3f8iNWxlLJMtXwYKyLtKJvOGXeW9feKSKKIJGZmZlZHJOXmfP386fanWfzc6hHanfgfh16+kr165SpVS1Wm6PcCTU+7H+FYVuEYEfEEgoFsx/0IYD5wpzFmR0U/wBjzljEm3hgTHxYWdmHPQKmzEA8PrhjxVzb3nUFISTYNZ/Ul8bU7yd6XbnU0papVZYp+DdBKRKJExBsYBiwsN2YhZR+2AgwBfjTGGBGpC3wFTDbG6GEQyhKxCYMoHLeS1WG30iHzSwL+E8/qtyZy7Mghq6MpVS3OW/SOOfeJwCIgBZhrjEkWkWdFZIBj2LtAPRFJBR4CTh6CORFoCTwlIkmOPw2c/iyUOo8GTZrRc+K7HLjzZzbUSSB+7yzk1Y6s/uCvFOQeszqeUlVKjIt9bTw+Pt4kJiZaHUO5ue2bVnH866folL+SLOqS3n4isQMewNNbDwpTNZOIrDXGxFe0Tr8Zq2qlVjHd6PToIjZfN5dDXuHEb/4HB5/rwPqv39bDMZXb0aJXtVr7Htdx+WM/s+6KtygQX+JWP8zOKZ1JXjpPT5Km3IYWvar1xMODTlffRuRf17Eq7gV8SvNot/RuUp7rzfbEH6yOp9Ql06JXysFms1eaboQAAAtcSURBVNFt4HjqT97AiujHCCvaQ6svB7PhX9ezO2WN1fGUumha9EqV4+PjR8/hk/H980ZWXPZHmueuJ2LONSS+8gcO7NpmdTylLpgWvVJnERgUTM+7nqN44npWNrqd9kd+JHRGD1ZPH8uRQ+W/M6iU69KiV+o8QsMa03PCdI7cs4p1oTfQ6eA8fF6PY9WMh8k9phcoV65Pi16pSmrctAXd/zSLjOFL2BLYnW6736b4pQ6s+uhZCgtyrY6n1Flp0St1gSLbxBL/l4VsHbCQ3T6t6Pbrixx5vgOJ8/9Nib3Y6nhK/Y4WvVIXqU2nPsRM/pFN/T7gmC2E+A1PkPHPOJK++xBTWmp1PKVO0aJX6hKICDG9B9Ly8dWs7TYNMaXErpjI9n92Y8uKL62OpxSgRa+UU3jYPOh8/SiaPJ7EqphnCbZn0/a7EWx6/ipSk5ZbHU/Vclr0SjmRp5c33Qb/ieBHN/G/lg/RtGAbLRfcxLoXbyFj+0ar46laSoteqSrg6xdAj5FPIw9u5H/hY2hzbAWNZvVh9bQ7SE1arnP4qlrpaYqVqgaZB3azY94zdMpcgLeUcIhQdta7Et/2NxLd/UZ8/QKsjqhquHOdpliLXqlqdPjQXlJ/mY9n6iLanFiFvxSSZ3zYFhiPvWV/WvS8ldCGEVbHVDWQFr1SLqggP5dfV35DXvKXRGUtoyHZlBrhV+82HI24msZdb6FZdCfEQ2dY1flp0Svl4kxpKTs2ryQzcT719/5Iq5JUAPZKQ/aE9SGww81Ed70OL70CljoLLXqlaphDe3eyc8Vn+KR9x+V56/CRYo7hz/ag7pjo62nVaxDBIWFWx1QuRIteqRos70QOW1d8gX3LV7Q4+gv1yMFuPNjmE8PxyGto2v1Wwpu3szqmspgWvVJuoqSkhO3rlnJk/X9pdGApUaW7AEj3aMqBRn2pGzuAVp36YvP0tDipqm5a9Eq5qb1pKexZ+RmBu74numATXlLCYeqwo24vbJffSHTPmwkIqmt1TFUNtOiVqgVyjmSxfcV82PYtrY/9jzrkUmi82OoXS0Hz64jseSsNI1pYHVNVES16pWqZ4qJCtq35nhMbvyDi0FIizAEAUm0tyGxyFfXibqbZ5fH6RS03okWvVC1mSkvZ/WsS+1bPp+6eH4guSsFDDCVG2O/RkGzfy8gPboktrDVBTdvRuGUswaF6RE9No0WvlDrl8KG97Ez8lqL9KXgf2U5IXjrhJXvxkd8umpJNMAe8m5Eb1JzSeq0JCG9Lg6gYGkS00C9wuSgteqXUOZXY7RzYvY2s9E3k70vBI+tX6uSm07h4F8H8dpnEPOPDPs8IjgY0pzi0Jb6N2hB6WQyNm7fD28fXwmegzlX0egyWUgqbpyfhzdv97nh8U1pKVuY+Du7YyImMZEzmr/gf30HEsfU0OvY9pAMrwW482OPRiCy/SArrtsTWIJrgZu1p3KIDQcGhljwn9RsteqXUWYmHB/UbRlC/YQRwwxnrco8fZd+OTRzdnYz94FZ8jqYSmp9Ok72r8N5XAkll4w4RyiGfsmkgwqIJbHI5DVt0pF6jZjoNVE106kYp5VTFRYXsT08hK30zhftTsB1OpW5uGk2K9xAo+afGHcePg7Ym5HmFUuRdF7tvKPjXxyOwHl51GuAbHEZASCPqhDSkTkgYHvolsHPSqRulVLXx8vahWetYmrWOPWO5KS3l0P7dHEzbQO7eFMjaht+JPfgXH6V+4S6Cc44RIAUVPmaJEY5IEMc86pDrWZdCr7oU+4RS6heK+NfDFhSGd50w/Os2JDC0IXXrN8bXP6g6nm6NoEWvlKoW4uFBg/BIGoRHAgMrHJOfe4Kcwwc5cfgAeUcPUXQ8E/uxTExeNh752XgWHMa3+Ah183cTlLuZuuYYnlLx1bryjTc5UocTtmDyvepS6B2K3Tek7LeGgHp4BYXhG9wA/5AGBATXx8cvEF//ALy8fECk6l4IC1Sq6EWkP/AqYAPeMcY8X269D/AB0BnIBm4zxqQ71j0G3A2UAA8YYxY5Lb1Syq34BQTiFxAITSv3Dd7SkhJycrLJyd5P7pFDFOQcovhYJiW5WZCbha3gCN5FR/AvPkJoYQbBOcfOmD6qiN14UIg3BeJDofhQLD4UefhQ7OFLiYcPdpsvJTY/jM2HUk8/jKcfeJX9ES8/PLz98fDxx+btj6dPAJ6+/nj6+OPtF4i3r3/ZG4pfID5+AXjYbM542c7rvEUvIjbgdeAaIANYIyILjTFbTht2N3DEGNNSRIYBLwC3iUhbYBjQDmgC/CAirY0xJc5+Ikqp2sfDZiM4tAHBoQ0qvU1Bfh7HDh/i2OH95B89RGHOIex5R6E4H1OcD8X5SHE+UlKAhz0fD3s+ttJCPEsK8LafIKAoG29TiLcpxIdCfE3RGd9BuBCFxosC8aYQH4rEh4wGCXSf8OZFPda5VGaPviuQaoxJAxCROZT93nV60Q8EnnHcnge8JiLiWD7HGFMI7BSRVMfj/c858ZVS6sL4+vnje2oKyTlK7HYK8nMpyD9BYd4JigvzKMrPpbggF3thHvbCXEqL8igtyqe0MA9jz4eifLDnIfaTbygFEBzutEynq0zRhwN7TrufAXQ72xhjjF1EcoB6juUry237u2ciIvcC9wI0a9asstmVUsol2Dw9CQgKJiAo2OooFXKJg1iNMW8ZY+KNMfFhYXqODaWUcqbKFP1eoOlp9yMcyyocIyKeQDBlH8pWZlullFJVqDJFvwZoJSJRIuJN2YerC8uNWQiMctweAvxoyr6JtRAYJiI+IhIFtAJWOye6UkqpyjjvHL1jzn0isIiywytnGGOSReRZINEYsxB4F/jQ8WHrYcreDHCMm0vZB7d24D494kYppaqXngJBKaXcwLlOgeASH8YqpZSqOlr0Sinl5rTolVLKzbncHL2IZAK7LuEh6gNZTopT0+lrcSZ9PX6jr8WZ3OH1uMwYU+EXkVyu6C+ViCSe7QOJ2kZfizPp6/EbfS3O5O6vh07dKKWUm9OiV0opN+eORf+W1QFciL4WZ9LX4zf6WpzJrV8Pt5ujV0opdSZ33KNXSil1Gi16pZRyc25T9CLSX0S2iUiqiEy2Oo+VRKSpiCwRkS0ikiwif7I6k9VExCYi60XkS6uzWE1E6orIPBHZKiIpItLD6kxWEpFJjn8nm0XkYxHxtTqTs7lF0Z92XdvrgbbAcMf1amsrO/BnY0xboDtwXy1/PQD+BKRYHcJFvAp8a4xpA3SkFr8uIhIOPADEG2PaU3aG3mHWpnI+tyh6TruurTGmCDh5XdtayRiz3xizznH7OGX/kKvmYpQ1gIhEADcC71idxWoiEgxcSdmpxTHGFBljjlqbynKegJ/jokn+wD6L8ziduxR9Rde1rbXFdjoRiQTigFXWJrHUK8AjQKnVQVxAFJAJvOeYynpHRAKsDmUVY8xeYCqwG9gP5BhjvrM2lfO5S9GrCohIIPAZ8KAx5pjVeawgIjcBh4wxa63O4iI8gU7AG8aYOCAXqLWfaYlICGW//UcBTYAAERlpbSrnc5ei12vTliMiXpSV/GxjzOdW57FQL2CAiKRTNqV3lYjMsjaSpTKADGPMyd/w5lFW/LXV1cBOY0ymMaYY+BzoaXEmp3OXoq/MdW1rDRERyuZgU4wxL1mdx0rGmMeMMRHGmEjK/r/40RjjdntslWWMOQDsEZFox6J+lF3qs7baDXQXEX/Hv5t+uOGH0+e9ZmxNcLbr2locy0q9gDuATSKS5Fj2uDHmawszKddxPzDbsVOUBtxlcR7LGGNWicg8YB1lR6utxw1Ph6CnQFBKKTfnLlM3SimlzkKLXiml3JwWvVJKuTkteqWUcnNa9Eop5ea06JVSys1p0SullJv7f4zELPWN0vANAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2l227HiFTZWF"
      },
      "source": [
        "## Time for prediction!"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VP5giYZ1tkWg"
      },
      "source": [
        "model.eval()\n",
        "prediction = model(X_val)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "81ZuTiguaWTN",
        "outputId": "aceb8bef-a300-4527-d84c-b117082294b8"
      },
      "source": [
        "predicted = plt.scatter(X_val.cpu(), prediction.detach().cpu())\n",
        "original = plt.scatter(X_train.cpu(), y_train.cpu())\n",
        "plt.legend((predicted, original), (\"Predicted\", \"Original\"));"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Qb1A0VT_8af8"
      },
      "source": [
        "### So, here we can see that our neural network was able to approximate sine function very well."
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bj0RhPCZTfc5"
      },
      "source": [
        "## State_dict of model and optimizer"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "oQfpMzNHL2F3",
        "outputId": "b34d98df-371e-469d-8152-a165ad6bbe98"
      },
      "source": [
        "print(\"Model's state_dict:\")\n",
        "for param_tensor in model.state_dict():\n",
        "    print(param_tensor, \"\\t\", model.state_dict()[param_tensor].size())\n",
        "\n",
        "# Print optimizer's state_dict\n",
        "print(\"Optimizer's state_dict:\")\n",
        "for var_name in optimizer.state_dict():\n",
        "    print(var_name, \"\\t\", optimizer.state_dict()[var_name])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model's state_dict:\n",
            "regressor.0.weight \t torch.Size([1024, 1])\n",
            "regressor.0.bias \t torch.Size([1024])\n",
            "regressor.2.weight \t torch.Size([1024, 1024])\n",
            "regressor.2.bias \t torch.Size([1024])\n",
            "regressor.4.weight \t torch.Size([1, 1024])\n",
            "regressor.4.bias \t torch.Size([1])\n",
            "Optimizer's state_dict:\n",
            "state \t {0: {'step': 15630, 'exp_avg': tensor([[ 4.6025e-05],\n",
            "        [ 7.3237e-09],\n",
            "        [-1.8306e-05],\n",
            "        ...,\n",
            "        [ 2.3937e-05],\n",
            "        [-3.3443e-06],\n",
            "        [ 6.3547e-06]], device='cuda:0'), 'exp_avg_sq': tensor([[2.7239e-07],\n",
            "        [5.5221e-15],\n",
            "        [5.6076e-08],\n",
            "        ...,\n",
            "        [3.0065e-08],\n",
            "        [1.8366e-08],\n",
            "        [2.7159e-08]], device='cuda:0')}, 1: {'step': 15630, 'exp_avg': tensor([-2.0342e-05,  8.6167e-08,  1.3242e-05,  ..., -1.3676e-05,\n",
            "         2.4177e-06, -2.2390e-07], device='cuda:0'), 'exp_avg_sq': tensor([2.8005e-08, 6.9960e-13, 2.2978e-09,  ..., 3.1504e-09, 1.0814e-09,\n",
            "        1.8065e-09], device='cuda:0')}, 2: {'step': 15630, 'exp_avg': tensor([[-9.5109e-06, -1.2803e-08,  5.3582e-06,  ..., -2.6647e-06,\n",
            "          1.5737e-05,  3.9083e-06],\n",
            "        [-5.6517e-06, -7.6079e-09, -3.2443e-06,  ...,  5.3052e-06,\n",
            "         -5.8846e-06,  3.7642e-07],\n",
            "        [ 1.0314e-05,  1.3884e-08,  5.9208e-06,  ..., -9.6804e-06,\n",
            "          1.0741e-05, -6.8595e-07],\n",
            "        ...,\n",
            "        [-1.7069e-05, -2.2927e-08,  8.5206e-06,  ..., -8.1828e-06,\n",
            "          2.1549e-05,  3.4623e-06],\n",
            "        [-9.4069e-06, -1.2663e-08, -5.4000e-06,  ...,  8.8297e-06,\n",
            "         -9.7949e-06,  6.2621e-07],\n",
            "        [ 7.8935e-06,  0.0000e+00,  4.6802e-06,  ..., -7.8983e-06,\n",
            "          8.4953e-06, -6.1740e-07]], device='cuda:0'), 'exp_avg_sq': tensor([[8.7415e-10, 1.3410e-14, 3.3180e-10,  ..., 9.0307e-09, 1.8338e-08,\n",
            "         6.7497e-09],\n",
            "        [3.1673e-10, 4.9371e-15, 1.4868e-09,  ..., 1.4235e-08, 5.3265e-08,\n",
            "         1.5084e-08],\n",
            "        [1.0730e-09, 1.6643e-14, 5.0193e-09,  ..., 4.8027e-08, 1.7976e-07,\n",
            "         5.0900e-08],\n",
            "        ...,\n",
            "        [2.9070e-09, 4.4087e-14, 8.1732e-10,  ..., 2.4582e-08, 4.6013e-08,\n",
            "         1.7539e-08],\n",
            "        [8.6079e-10, 1.3500e-14, 4.0608e-09,  ..., 3.8907e-08, 1.4553e-07,\n",
            "         4.1221e-08],\n",
            "        [5.4460e-10, 0.0000e+00, 3.0548e-09,  ..., 2.8733e-08, 1.0944e-07,\n",
            "         3.0818e-08]], device='cuda:0')}, 3: {'step': 15630, 'exp_avg': tensor([-6.5970e-06,  6.7459e-06, -1.2310e-05,  ..., -1.3571e-05,\n",
            "         1.1228e-05, -9.9987e-06], device='cuda:0'), 'exp_avg_sq': tensor([3.9559e-09, 4.0460e-09, 1.3649e-08,  ..., 1.1347e-08, 1.1059e-08,\n",
            "        7.9782e-09], device='cuda:0')}, 4: {'step': 15630, 'exp_avg': tensor([[-0.0007, -0.0002,  0.0001,  ..., -0.0007, -0.0005,  0.0003]],\n",
            "       device='cuda:0'), 'exp_avg_sq': tensor([[4.5959e-06, 8.5489e-06, 4.6649e-05,  ..., 4.3342e-06, 5.8228e-06,\n",
            "         4.0915e-06]], device='cuda:0')}, 5: {'step': 15630, 'exp_avg': tensor([0.0005], device='cuda:0'), 'exp_avg_sq': tensor([2.4550e-05], device='cuda:0')}}\n",
            "param_groups \t [{'lr': 1e-06, 'betas': (0.9, 0.999), 'eps': 1e-08, 'weight_decay': 0, 'amsgrad': False, 'params': [0, 1, 2, 3, 4, 5]}]\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "70xKgSTyTlgh"
      },
      "source": [
        "## Saving model using save() and state_dict()"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L_zde1DiNSue"
      },
      "source": [
        "PATH = '/content/sine.pt'\n",
        "torch.save(model.state_dict(), PATH)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3FYvrYEYTppk"
      },
      "source": [
        "## Loading model using load() and load_state_dict()"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bwVb4kb_N7Q2",
        "outputId": "a78f572c-a62a-429e-8238-fb0fd900a8e9"
      },
      "source": [
        "device = torch.device(\"cuda\")\n",
        "model = SineApproximator()\n",
        "model.load_state_dict(torch.load(PATH))\n",
        "model.to(device)\n",
        "model.eval()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "SineApproximator(\n",
              "  (regressor): Sequential(\n",
              "    (0): Linear(in_features=1, out_features=1024, bias=True)\n",
              "    (1): ReLU(inplace=True)\n",
              "    (2): Linear(in_features=1024, out_features=1024, bias=True)\n",
              "    (3): ReLU(inplace=True)\n",
              "    (4): Linear(in_features=1024, out_features=1, bias=True)\n",
              "  )\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 49
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "3DHdz13eTtl8"
      },
      "source": [
        "## Trying pre-trained model for inference/training"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 265
        },
        "id": "jpBrFM3qOKzC",
        "outputId": "d5feedde-0f0c-45cb-a5c9-7fafec335a73"
      },
      "source": [
        "new_pred = model(X_val)\n",
        "pred = plt.scatter(X_val.cpu(), new_pred.detach().cpu())\n",
        "original = plt.scatter(X_train.cpu(), y_train.cpu())\n",
        "plt.legend((pred, original), (\"Predicted\", \"Original\"));"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}