COSMOS_ML_03_Multilayer_Perceptron.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/sungjuGit/COSMOS_Ju/blob/main/COSMOS_ML_03_Multilayer_Perceptron.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "---\n",
        "### Set Things Up"
      ],
      "metadata": {
        "id": "GwKUFx77YD4N"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "J3VsmcZU4O8I"
      },
      "source": [
        "# 1 - Multilayer Perceptron\n",
        "\n",
        "We build machine learning models (specifically, neural networks) to perform image classification using PyTorch.\n",
        "\n",
        "We'll start with one of the most basic neural network architectures, a multilayer perceptron (MLP), also known as a feedforward network. The dataset we'll be using is the famous MNIST dataset, a dataset of 28x28 black and white images consisting of handwritten digits, 0 to 9.\n",
        "\n",
        "![](https://github.com/bentrevett/pytorch-image-classification/blob/master/assets/mlp-mnist.png?raw=1)\n",
        "\n",
        "We'll process the dataset, build our model, and then train our model.\n",
        "\n",
        "\n",
        "This notebook was adapted from (https://github.com/bentrevett/pytorch-image-classification)."
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "---\n",
        "## Set Things Up"
      ],
      "metadata": {
        "id": "fG0bszF0YOf7"
      }
    },
    {
      "cell_type": "markdown",
      "source": [
        "Let's start by importing all the modules we'll need. The main ones we need to import are:\n",
        "- torch for general PyTorch functionality\n",
        "- torch.nn and torch.nn.functional for neural network based functions\n",
        "- torch.optim for our optimizer which will update the parameters of our neural network\n",
        "- torch.utils.data for handling the dataset\n",
        "- torchvision.transforms for data augmentation\n",
        "- torchvision.datasets for loading the dataset\n",
        "- matplotlib for plotting"
      ],
      "metadata": {
        "id": "b-JfdsGDYSZ6"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "CHadwpvmREvE",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "# @title Import Libraries\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "import torch.utils.data as data\n",
        "\n",
        "import torchvision.transforms as transforms\n",
        "import torchvision.datasets as datasets\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "import time\n",
        "import random\n",
        "\n",
        "SEED = 1234\n",
        "\n",
        "torch.manual_seed(SEED)\n",
        "torch.cuda.manual_seed(SEED)\n",
        "torch.backends.cudnn.deterministic = True"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yDMpvXwB4O8Q"
      },
      "source": [
        "The next thing we'll do is load our dataset. This will automatically download the training set for the MNIST dataset and save it in a folder called `.data`. It will create the folder if it does not exist."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "UOIQVy-cREvQ",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "2853065b-f43d-4f0d-d40f-513e1de8483e"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-images-idx3-ubyte.gz to .data/MNIST/raw/train-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 9912422/9912422 [00:00<00:00, 137559510.20it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting .data/MNIST/raw/train-images-idx3-ubyte.gz to .data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/train-labels-idx1-ubyte.gz to .data/MNIST/raw/train-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 28881/28881 [00:00<00:00, 116364739.50it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting .data/MNIST/raw/train-labels-idx1-ubyte.gz to .data/MNIST/raw\n",
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-images-idx3-ubyte.gz to .data/MNIST/raw/t10k-images-idx3-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 1648877/1648877 [00:00<00:00, 87289898.86it/s]"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting .data/MNIST/raw/t10k-images-idx3-ubyte.gz to .data/MNIST/raw\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz\n",
            "Downloading http://yann.lecun.com/exdb/mnist/t10k-labels-idx1-ubyte.gz to .data/MNIST/raw/t10k-labels-idx1-ubyte.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 4542/4542 [00:00<00:00, 2381911.57it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting .data/MNIST/raw/t10k-labels-idx1-ubyte.gz to .data/MNIST/raw\n",
            "\n"
          ]
        }
      ],
      "source": [
        "ROOT = '.data'\n",
        "\n",
        "train_data = datasets.MNIST(root=ROOT,\n",
        "                            train=True,\n",
        "                            download=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gMThB8Vs4O8V"
      },
      "source": [
        "Next, we want to *normalize* our data. This means we want it to have a mean of zero and a standard deviation of one. Why do we want to do this? Normalizing our data allows our models to train faster and to also help them avoid local minima, i.e. train more reliably. The image data has values between 0-255, which we want to scale between 0-1, so we divide by 255."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "FhM076c7Zzrv"
      },
      "outputs": [],
      "source": [
        "mean = train_data.data.float().mean() / 255\n",
        "std = train_data.data.float().std() / 255"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U50VOMAW4O8d"
      },
      "source": [
        "Now we've calculated our mean and standard deviation how do we actually use them? We use Torchvision's `transforms`.\n",
        "\n",
        "A `transform` states how our data should be augmented and processed. Data augmentation involves manipulating the available training data in a way that artificially creates more training examples. We use `transforms.Compose` to built a list of transformations that will be applied to the image."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "WuYq4cRdREvV",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "# @title Transform data\n",
        "train_transforms = transforms.Compose([\n",
        "                            transforms.RandomRotation(5, fill=(0,)),\n",
        "                            transforms.RandomCrop(28, padding=2),\n",
        "                            transforms.ToTensor(),\n",
        "                            transforms.Normalize(mean=[mean], std=[std])\n",
        "                                      ])\n",
        "\n",
        "test_transforms = transforms.Compose([\n",
        "                           transforms.ToTensor(),\n",
        "                           transforms.Normalize(mean=[mean], std=[std])\n",
        "                                     ])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "73S5zyox4O8h"
      },
      "source": [
        "Now we have defined our transforms we can then load the train and test data with the relevant transforms defined."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "GPvVD4VLREva"
      },
      "outputs": [],
      "source": [
        "train_data = datasets.MNIST(root=ROOT,\n",
        "                            train=True,\n",
        "                            download=True,\n",
        "                            transform=train_transforms)\n",
        "\n",
        "test_data = datasets.MNIST(root=ROOT,\n",
        "                           train=False,\n",
        "                           download=True,\n",
        "                           transform=test_transforms)"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "---\n",
        "## Check Out Data"
      ],
      "metadata": {
        "id": "wPWbLtT5YtNR"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ayyHLZjg4O8o"
      },
      "source": [
        "We can simply check the `len` of the datasets to see how many examples are within each."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fPo3pfgciRo3",
        "outputId": "88dabffa-8f23-4cc4-f4c0-4f7a5e2becd4"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of training examples: 60000\n",
            "Number of testing examples: 10000\n"
          ]
        }
      ],
      "source": [
        "print(f'Number of training examples: {len(train_data)}')\n",
        "print(f'Number of testing examples: {len(test_data)}')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2a23Msn44O8u"
      },
      "source": [
        "We can get a look at some of the images within our dataset to see what we're working with. The function below plots a square grid of images. If you supply less than a complete square number of images it will ignore the last few."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "v2IUyQZcaWh6"
      },
      "outputs": [],
      "source": [
        "# @title Plot Images function\n",
        "def plot_images(images):\n",
        "\n",
        "    n_images = len(images)\n",
        "\n",
        "    rows = int(np.sqrt(n_images))\n",
        "    cols = int(np.sqrt(n_images))\n",
        "\n",
        "    fig = plt.figure()\n",
        "    for i in range(rows*cols):\n",
        "        ax = fig.add_subplot(rows, cols, i+1)\n",
        "        ax.imshow(images[i].view(28, 28).cpu().numpy(), cmap='bone')\n",
        "        ax.axis('off')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "M6ZSTtn04O80"
      },
      "source": [
        "Let's load 25 images. These will have been processed through our transforms, so will be randomly rotated and cropped.\n",
        "\n",
        "It's a good practice to see your data with your transforms applied, so you can ensure they look sensible. For example, it wouldn't make sense to flip the digits horizontally or vertically unless you are expecting to see what in your test data."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 406
        },
        "id": "sxEwZKjWROlm",
        "outputId": "de34ced9-b396-4878-9086-535514e7a463"
      },
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 25 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ],
      "source": [
        "N_IMAGES = 25\n",
        "\n",
        "images = [image for image, label in [train_data[i] for i in range(N_IMAGES)]]\n",
        "\n",
        "plot_images(images)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "C7vOeo1a4O9W"
      },
      "source": [
        "Next, we'll define a `DataLoader` for each of the training and test sets. We can iterate over these, and they will yield batches of images and labels which we can use to train our model.\n",
        "\n",
        "We only need to shuffle our training set as it will be used for stochastic gradient descent, and we want each batch to be different between epochs.\n",
        "\n",
        "Ideally, we want to use the biggest batch size that we can. The 64 here is relatively small and can be increased if our hardware can handle it."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "sGdKSORpREvr"
      },
      "outputs": [],
      "source": [
        "BATCH_SIZE = 64\n",
        "\n",
        "train_loader = data.DataLoader(train_data,\n",
        "                                 shuffle=True,\n",
        "                                 batch_size=BATCH_SIZE)\n",
        "test_loader = data.DataLoader(test_data,\n",
        "                                batch_size=BATCH_SIZE)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KZuPIyvN4O9b"
      },
      "source": [
        "---\n",
        "## Define the Model\n",
        "\n"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Our model will be a neural network, specifically a multilayer perceptron (MLP) with two hidden layers. The image below shows the archicture of the model.\n",
        "\n",
        "![](https://github.com/bentrevett/pytorch-image-classification/blob/master/assets/mlp-mnist.png?raw=1)\n",
        "\n",
        "Specifically, first we will flatten our 1x28x28 (1 color channel, 28 pixels height and width) image into a 784 element vector, also called 784 *features*. We flatten our input, as MLPs cannot handle two or three-dimensional data. Next, the 784 dimensional input is passed through the first hidden layer to transform it into 250 dimensions. Then, another hidden layer, which will transform it to 100 dimensions. Finally, an output layer which will transform it into a 10 dimensional vector. The output dimension should equal the number of classes within your data. Here we have ten digits, 0 - 9, so need our output to be 10 dimensions.\n",
        "\n",
        "The transformation between 784 to 250, 250 to 100 and 100 to 10 dimensions are done by `Linear` layers. These are also known as fully connected or affine layers. In these layers, every element in one layer is connected to every element in the next. We can think of these elements as *neurons*, as this architecture is inspired by how the human brain is made of millions of interconnected nodes, also called neurons.\n",
        "\n",
        "Each connection between a neuron in one layer and a neuron in the next has a *weight* associated with it. The input to one neuron is the sum of the weighted values of all neurons in the previous layer connected to it, plus a bias term. The neuron then applies an *activation function* to this weighted sum. This activation function is a non-linear function that allows the neural network to learn non-linear functions between inputs and outputs.\n",
        "\n",
        "Why did we choose hidden dimensions of 250 and 100 elements? Why did we only have two hidden layers? There is no magic formula to tell us how many layers to use and how many neurons to have in each layer, and there is most probably a better set of values. However, the general idea is that neural networks extract features from data. Layers closer to the input learn to extract general features (e.g. lines, curves, edges), whilst later layers combine the features extracted from the previous layer into more high level features (e.g. the intersection of two lines making a cross, multiple curves make a circle). We force our neural network to learn these features by reducing the number of neurons in each layer. This way, it has to learn to compress information by extracting only the useful and general features. Thus, we want a neural network with multiple layers and some sort of information compression (reduced number of neurons in subsequent layers)."
      ],
      "metadata": {
        "id": "Y57sZWLQZUmq"
      }
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "lAqzcW9XREvu"
      },
      "outputs": [],
      "source": [
        "INPUT_DIM = 28 * 28\n",
        "OUTPUT_DIM = 10\n",
        "\n",
        "model = nn.Sequential(\n",
        "          nn.Flatten(),\n",
        "          nn.Linear(INPUT_DIM,250),\n",
        "          nn.ReLU(),\n",
        "          nn.Linear(250,100),\n",
        "          nn.ReLU(),\n",
        "          nn.Linear(100,OUTPUT_DIM)\n",
        "        )"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "F3bacWHd4O9j"
      },
      "source": [
        "Let's calculate the number of trainable parameters (weights and biases) in our model."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "1gF2fmTBREv8",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "cellView": "form",
        "outputId": "d13eed38-ea2c-4f84-8cb9-5e2d748f121c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "The model has 222,360 trainable parameters\n"
          ]
        }
      ],
      "source": [
        "# @title Number of parameters\n",
        "def count_parameters(model):\n",
        "    return sum(p.numel() for p in model.parameters() if p.requires_grad)\n",
        "\n",
        "print(f'The model has {count_parameters(model):,} trainable parameters')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gik6Q5ZB4O9q"
      },
      "source": [
        "The first layer has 784 neurons connected to 250 neurons, so 784*250 weighted connections plus 250 bias terms.\n",
        "\n",
        "The second layer has 250 neurons connected to 100 neurons, 250*100 weighted connections plus 100 bias terms.\n",
        "\n",
        "The third layer has 100 neurons connected to 10 neurons, 100*10 weighted connections plus 10 bias terms.\n",
        "\n",
        "$$784 \\cdot 250 + 250 + 250 \\cdot 100 + 100 + 100 \\cdot 10 + 10 = 222,360 $$"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IiL9hZqX4O9x"
      },
      "source": [
        "---\n",
        "## Train and Test the Model\n",
        "\n",
        "- pass a batch of data through your model\n",
        "- calculate the loss of your batch by comparing your model's predictions against the actual labels\n",
        "- calculate the gradient of each of your parameters with respect to the loss\n",
        "- update each of your parameters by subtracting their gradient multiplied by a small *learning rate* parameter\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "eC2P8WJuREv_"
      },
      "outputs": [],
      "source": [
        "optimizer = optim.Adam(model.parameters())"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "gBvdk2RR4O95"
      },
      "source": [
        "We use as our loss `CrossEntropyLoss`"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "OO2BL-qBREwD"
      },
      "outputs": [],
      "source": [
        "loss_fn = nn.CrossEntropyLoss()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "7QgyFD6gREwH",
        "cellView": "form"
      },
      "outputs": [],
      "source": [
        "# @title Use GPU if available\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')\n",
        "\n",
        "model = model.to(device)\n",
        "loss_fn = loss_fn.to(device)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Bd0F_7i_4O-M"
      },
      "source": [
        "Next, we'll define a function to calculate the accuracy of our model. This takes the index of the highest value for your prediction and compares it against the actual class label. We then divide how many our model got correct by the amount in the batch to calculate accuracy across the batch."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VkttGrmkREwP"
      },
      "outputs": [],
      "source": [
        "# @title Function to calculate accuracy\n",
        "def calculate_accuracy(y_pred, y):\n",
        "    top_pred = y_pred.argmax(1, keepdim=True)\n",
        "    correct = top_pred.eq(y.view_as(top_pred)).sum()\n",
        "    acc = correct.float() / y.shape[0]\n",
        "    return acc"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "0kCtKBMfREwS"
      },
      "outputs": [],
      "source": [
        "def train(model, loader, optimizer, loss_fn, device):\n",
        "\n",
        "    epoch_loss = 0\n",
        "    epoch_acc = 0\n",
        "\n",
        "    model.train()\n",
        "\n",
        "    for (x, y) in loader:\n",
        "\n",
        "        x = x.to(device)\n",
        "        y = y.to(device)\n",
        "\n",
        "        optimizer.zero_grad()\n",
        "\n",
        "        y_pred = model(x)\n",
        "\n",
        "        loss = loss_fn(y_pred, y)\n",
        "\n",
        "        acc = calculate_accuracy(y_pred, y)\n",
        "\n",
        "        loss.backward()\n",
        "\n",
        "        optimizer.step()\n",
        "\n",
        "        epoch_loss += loss.item()\n",
        "        epoch_acc += acc.item()\n",
        "\n",
        "    return epoch_loss / len(loader), epoch_acc / len(loader)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dySDirym4O-c"
      },
      "source": [
        "The final step before training is to define a small function to tell us how long an epoch took."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "ENrsvnEZREwZ"
      },
      "outputs": [],
      "source": [
        "def epoch_time(start_time, end_time):\n",
        "    elapsed_time = end_time - start_time\n",
        "    elapsed_mins = int(elapsed_time / 60)\n",
        "    elapsed_secs = int(elapsed_time - (elapsed_mins * 60))\n",
        "    return elapsed_mins, elapsed_secs"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uHn-rDvA4O-h"
      },
      "source": [
        "We're finally ready to train!\n",
        "\n",
        "During each epoch we calculate the training loss and accuracy, followed by the validation loss and accuracy. We then check if the validation loss achieved is the best validation loss we have seen. If so, we save our model's parameters (called a `state_dict`)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p-gtfzafREwc",
        "outputId": "f65ae362-41db-408d-cfb1-9143dbadc568"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch: 01 | Epoch Time: 0m 26s\n",
            "\tTrain Loss: 0.380 | Train Acc: 88.23%\n",
            "Epoch: 02 | Epoch Time: 0m 27s\n",
            "\tTrain Loss: 0.163 | Train Acc: 95.02%\n",
            "Epoch: 03 | Epoch Time: 0m 26s\n",
            "\tTrain Loss: 0.128 | Train Acc: 96.06%\n",
            "Epoch: 04 | Epoch Time: 0m 26s\n",
            "\tTrain Loss: 0.115 | Train Acc: 96.45%\n",
            "Epoch: 05 | Epoch Time: 0m 26s\n",
            "\tTrain Loss: 0.101 | Train Acc: 96.72%\n",
            "Epoch: 06 | Epoch Time: 0m 26s\n",
            "\tTrain Loss: 0.094 | Train Acc: 97.07%\n",
            "Epoch: 07 | Epoch Time: 0m 30s\n",
            "\tTrain Loss: 0.086 | Train Acc: 97.33%\n",
            "Epoch: 08 | Epoch Time: 0m 31s\n",
            "\tTrain Loss: 0.083 | Train Acc: 97.44%\n",
            "Epoch: 09 | Epoch Time: 0m 28s\n",
            "\tTrain Loss: 0.080 | Train Acc: 97.47%\n",
            "Epoch: 10 | Epoch Time: 0m 27s\n",
            "\tTrain Loss: 0.074 | Train Acc: 97.70%\n"
          ]
        }
      ],
      "source": [
        "EPOCHS = 10\n",
        "\n",
        "for epoch in range(EPOCHS):\n",
        "\n",
        "    start_time = time.monotonic()\n",
        "\n",
        "    train_loss, train_acc = train(model, train_loader, optimizer, loss_fn, device)\n",
        "\n",
        "    end_time = time.monotonic()\n",
        "\n",
        "    epoch_mins, epoch_secs = epoch_time(start_time, end_time)\n",
        "\n",
        "    print(f'Epoch: {epoch+1:02} | Epoch Time: {epoch_mins}m {epoch_secs}s')\n",
        "    print(f'\\tTrain Loss: {train_loss:.3f} | Train Acc: {train_acc*100:.2f}%')"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "test_idx = random.randint(0,10000-1) # out of 10000\n",
        "\n",
        "model.eval()\n",
        "\n",
        "with torch.no_grad():\n",
        "  test_pred = model(test_data[test_idx][0].unsqueeze_(0).to(device))\n",
        "  prob = torch.softmax(test_pred, dim=-1)\n",
        "  print(torch.argmax(prob, 1))\n",
        "\n",
        "plt.imshow(test_data[test_idx][0].view(28, 28).cpu().numpy(), cmap='bone')\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 466
        },
        "id": "w8UEQ1X9onpr",
        "outputId": "1bf6f5cd-676f-45f4-a8eb-3e0827d5f1d7"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "tensor([8], device='cuda:0')\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<matplotlib.image.AxesImage at 0x7a298e617280>"
            ]
          },
          "metadata": {},
          "execution_count": 19
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "9kuF4-Cx4O-X"
      },
      "source": [
        "The evaluation loop is similar to the training loop. The differences are:\n",
        "- we put our model into evaluation mode with `model.eval()`\n",
        "- we wrap the iterations inside a `with torch.no_grad()`\n",
        "- we do not zero gradients as we are not calculating any\n",
        "- we do not calculate gradients as we are not updating parameters\n",
        "- we do not take an optimizer step as we are not calculating gradients\n",
        "\n",
        "`torch.no_grad()` ensures that gradients are not calculated for whatever is inside the `with` block. As our model will not have to calculate gradients, it will be faster and use less memory."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "3XL4sRI8REwV"
      },
      "outputs": [],
      "source": [
        "def evaluate(model, loader, loss_fn, device):\n",
        "\n",
        "    epoch_loss = 0\n",
        "    epoch_acc = 0\n",
        "\n",
        "    model.eval()\n",
        "\n",
        "    with torch.no_grad():\n",
        "\n",
        "        for (x, y) in loader:\n",
        "\n",
        "            x = x.to(device)\n",
        "            y = y.to(device)\n",
        "\n",
        "            y_pred = model(x)\n",
        "\n",
        "            loss = loss_fn(y_pred, y)\n",
        "\n",
        "            acc = calculate_accuracy(y_pred, y)\n",
        "\n",
        "            epoch_loss += loss.item()\n",
        "            epoch_acc += acc.item()\n",
        "\n",
        "    return epoch_loss / len(loader), epoch_acc / len(loader)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {
        "id": "VGDeNGLhREwf",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "e5e76b2a-2cb6-4340-878f-8ad4062d8316"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Test Loss: 0.050 | Test Acc: 98.27%\n"
          ]
        }
      ],
      "source": [
        "test_loss, test_acc = evaluate(model, test_loader, loss_fn, device)\n",
        "print(f'Test Loss: {test_loss:.3f} | Test Acc: {test_acc*100:.2f}%')"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "machine_shape": "hm",
      "provenance": [],
      "collapsed_sections": [
        "fG0bszF0YOf7",
        "wPWbLtT5YtNR",
        "KZuPIyvN4O9b",
        "IiL9hZqX4O9x"
      ],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.8.3"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}