linear_regression.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "linear-regression.ipynb",
      "provenance": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JqOmUM8ZlEeZ"
      },
      "source": [
        "# Linear Regression"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 456
        },
        "id": "MRV78oqrgMeu",
        "outputId": "7dda9f24-dc9a-4e9d-80aa-e556050b5c19"
      },
      "source": [
        "# code by Jinshun Lin\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Hyper-parameters\n",
        "input_size = 1\n",
        "output_size = 1\n",
        "num_epochs = 59\n",
        "learning_rate = 0.001\n",
        "# Toy dataset\n",
        "x_train = np.array([[3.3], [4.4], [5.5], [6.71], [6.93], [4.168],\n",
        "                    [9.779], [6.182], [7.59], [2.167], [7.042],\n",
        "                    [10.791], [5.313], [7.997], [3.1]], dtype=np.float32)\n",
        "y_train = np.array([[1.7], [2.76], [2.09], [3.19], [1.694], [1.573], \n",
        "                    [3.366], [2.596], [2.53], [1.221], [2.827], \n",
        "                    [3.465], [1.65], [2.904], [1.3]], dtype=np.float32)\n",
        "\n",
        "# Linear regression model\n",
        "model = nn.Linear(input_size, output_size)\n",
        "\n",
        "# Loss and optimizer\n",
        "criterion = nn.MSELoss()\n",
        "optimizer = torch.optim.SGD(model.parameters(), lr=learning_rate)\n",
        "\n",
        "# Train the model\n",
        "for epoch in range(num_epochs):\n",
        "    # Convert numpy arrays to torch tensors\n",
        "    inputs = torch.from_numpy(x_train)\n",
        "    targets = torch.from_numpy(y_train)\n",
        "\n",
        "    # Forward pass\n",
        "    outputs = model(inputs)\n",
        "    loss = criterion(outputs, targets)\n",
        "\n",
        "    # Backward and optimize\n",
        "    optimizer.zero_grad()\n",
        "    loss.backward()\n",
        "    optimizer.step()\n",
        "\n",
        "    if (epoch+1) % 5 == 0:\n",
        "        print('Epoch [{}/{}], Loss: {:.4f}'.format(epoch+1, num_epochs, loss.item()))\n",
        "\n",
        "# Plot the graph\n",
        "predicted = model(torch.from_numpy(x_train)).detach().numpy()\n",
        "plt.plot(x_train, y_train, 'ro', label='Original data')\n",
        "plt.plot(x_train, predicted, label='Fitted line')\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "# Save the model checkpoint\n",
        "torch.save(model.state_dict(), 'model.ckpt')"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch [5/59], Loss: 1.5825\n",
            "Epoch [10/59], Loss: 0.9537\n",
            "Epoch [15/59], Loss: 0.6984\n",
            "Epoch [20/59], Loss: 0.5944\n",
            "Epoch [25/59], Loss: 0.5518\n",
            "Epoch [30/59], Loss: 0.5339\n",
            "Epoch [35/59], Loss: 0.5262\n",
            "Epoch [40/59], Loss: 0.5225\n",
            "Epoch [45/59], Loss: 0.5205\n",
            "Epoch [50/59], Loss: 0.5191\n",
            "Epoch [55/59], Loss: 0.5180\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}