Save_and_Load_models.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Save and Load models.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Vishwaak/Time-for-Machine/blob/master/Save_and_Load_models.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BLmoqpJ7NKap",
        "colab_type": "code",
        "outputId": "6ff6899b-e962-445c-85ce-c9d08d438956",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 280
        }
      },
      "source": [
        "import torch\n",
        "from torchvision import datasets, transforms\n",
        "import helper\n",
        "\n",
        "# Define a transform to normalize the data\n",
        "transform = transforms.Compose([transforms.ToTensor(),\n",
        "                                transforms.Normalize((0.5,), (0.5,))])\n",
        "# Download and load the training data\n",
        "trainset = datasets.FashionMNIST('~/.pytorch/F_MNIST_data/', download=True, train=True, transform=transform)\n",
        "trainloader = torch.utils.data.DataLoader(trainset, batch_size=64, shuffle=True)\n",
        "\n",
        "# Download and load the test data\n",
        "testset = datasets.FashionMNIST('~/.pytorch/F_MNIST_data/', download=True, train=False, transform=transform)\n",
        "testloader = torch.utils.data.DataLoader(testset, batch_size=64, shuffle=True)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "  0%|          | 114688/26421880 [00:00<00:22, 1146225.22it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-images-idx3-ubyte.gz to /root/.pytorch/F_MNIST_data/FashionMNIST/raw/train-images-idx3-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "26427392it [00:00, 81392483.16it/s]                             \n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Extracting /root/.pytorch/F_MNIST_data/FashionMNIST/raw/train-images-idx3-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "32768it [00:00, 552087.83it/s]\n",
            " 10%|▉         | 434176/4422102 [00:00<00:00, 4340977.56it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/train-labels-idx1-ubyte.gz to /root/.pytorch/F_MNIST_data/FashionMNIST/raw/train-labels-idx1-ubyte.gz\n",
            "Extracting /root/.pytorch/F_MNIST_data/FashionMNIST/raw/train-labels-idx1-ubyte.gz\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-images-idx3-ubyte.gz to /root/.pytorch/F_MNIST_data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "4423680it [00:00, 30246889.55it/s]                           \n",
            "8192it [00:00, 186294.25it/s]\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Extracting /root/.pytorch/F_MNIST_data/FashionMNIST/raw/t10k-images-idx3-ubyte.gz\n",
            "Downloading http://fashion-mnist.s3-website.eu-central-1.amazonaws.com/t10k-labels-idx1-ubyte.gz to /root/.pytorch/F_MNIST_data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz\n",
            "Extracting /root/.pytorch/F_MNIST_data/FashionMNIST/raw/t10k-labels-idx1-ubyte.gz\n",
            "Processing...\n",
            "Done!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TQXN7vIHNMXV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "image, label = next(iter(trainloader))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "i9Ed-kpaNPOs",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from torch import nn, optim\n",
        "import torch.nn.functional as F"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QTOApshMNY8J",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# TODO: Define your network architecture here\n",
        "class Classifier(nn.Module):\n",
        "    def __init__(self):\n",
        "        super().__init__()\n",
        "        self.fc1 = nn.Linear(784, 256)\n",
        "        self.fc2 = nn.Linear(256, 128)\n",
        "        self.fc3 = nn.Linear(128, 64)\n",
        "        self.fc4 = nn.Linear(64, 10)\n",
        "        \n",
        "    def forward(self, x):\n",
        "        # make sure input tensor is flattened\n",
        "        x = x.view(x.shape[0], -1)\n",
        "        \n",
        "        x = F.relu(self.fc1(x))\n",
        "        x = F.relu(self.fc2(x))\n",
        "        x = F.relu(self.fc3(x))\n",
        "        x = F.log_softmax(self.fc4(x), dim=1)\n",
        "        \n",
        "        return x"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PMeZ-kG3Na70",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = Classifier()\n",
        "criterion = nn.NLLLoss()\n",
        "optimizer = optim.Adam(model.parameters(), lr=0.003)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nv2COcPwNd0k",
        "colab_type": "code",
        "outputId": "78d4f733-9e1b-4ba8-b213-5f996a95df44",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        }
      },
      "source": [
        "epochs = 5\n",
        "\n",
        "for e in range(epochs):\n",
        "    running_loss = 0\n",
        "    for images, labels in trainloader:\n",
        "        log_ps = model(images)\n",
        "        loss = criterion(log_ps, labels)\n",
        "        \n",
        "        optimizer.zero_grad()\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "        \n",
        "        running_loss += loss.item()\n",
        "    else:\n",
        "        print(f\"Training loss: {running_loss/len(trainloader)}\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Training loss: 0.5189460875954963\n",
            "Training loss: 0.3930190469918729\n",
            "Training loss: 0.3554112941725676\n",
            "Training loss: 0.3307421572530257\n",
            "Training loss: 0.31786281673479944\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BfqTEAMnNf2g",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def view_classify(img, ps, version=\"MNIST\"):\n",
        "    ''' Function for viewing an image and it's predicted classes.\n",
        "    '''\n",
        "    ps = ps.data.numpy().squeeze()\n",
        "\n",
        "    fig, (ax1, ax2) = plt.subplots(figsize=(6,9), ncols=2)\n",
        "    ax1.imshow(img.resize_(1, 28, 28).numpy().squeeze())\n",
        "    ax1.axis('off')\n",
        "    ax2.barh(np.arange(10), ps)\n",
        "    ax2.set_aspect(0.1)\n",
        "    ax2.set_yticks(np.arange(10))\n",
        "    if version == \"MNIST\":\n",
        "        ax2.set_yticklabels(np.arange(10))\n",
        "    elif version == \"Fashion\":\n",
        "        ax2.set_yticklabels(['T-shirt/top',\n",
        "                            'Trouser',\n",
        "                            'Pullover',\n",
        "                            'Dress',\n",
        "                            'Coat',\n",
        "                            'Sandal',\n",
        "                            'Shirt',\n",
        "                            'Sneaker',\n",
        "                            'Bag',\n",
        "                            'Ankle Boot'], size='small');\n",
        "    ax2.set_title('Class Probability')\n",
        "    ax2.set_xlim(0, 1.1)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jm2Q1GbjN6NE",
        "colab_type": "code",
        "outputId": "846881c4-559e-4863-f9d8-542f18a4a595",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 214
        }
      },
      "source": [
        "%matplotlib inline\n",
        "%config InlineBackend.figure_format = 'retina'\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "dataiter = iter(testloader)\n",
        "images , labels = dataiter.next()\n",
        "img = images[1]\n",
        "img = img.view(img.shape[0] , -1)\n",
        "ps = torch.exp(model(img))\n",
        "\n",
        "view_classify(img , ps , version='Fashion')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuUAAAGKCAYAAACrcD/sAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMymXJEmSamZSLkmSJNXs/wHJOD5qWyMogwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x648 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "image/png": {
              "width": 370,
              "height": 197
            }
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VOu3z9IxpdcA",
        "colab_type": "code",
        "outputId": "7268ae1d-5811-482d-c5dd-f0bad9a2e387",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        }
      },
      "source": [
        "#code for fc_model\n",
        "\n",
        "class Network(nn.Module):\n",
        "    def __init__(self, input_size, output_size, hidden_layers, drop_p=0.5):\n",
        "        ''' Builds a feedforward network with arbitrary hidden layers.\n",
        "        \n",
        "            Arguments\n",
        "            ---------\n",
        "            input_size: integer, size of the input layer\n",
        "            output_size: integer, size of the output layer\n",
        "            hidden_layers: list of integers, the sizes of the hidden layers\n",
        "        \n",
        "        '''\n",
        "        super().__init__()\n",
        "        # Input to a hidden layer\n",
        "        self.hidden_layers = nn.ModuleList([nn.Linear(input_size, hidden_layers[0])])\n",
        "        \n",
        "        # Add a variable number of more hidden layers\n",
        "        layer_sizes = zip(hidden_layers[:-1], hidden_layers[1:])\n",
        "        self.hidden_layers.extend([nn.Linear(h1, h2) for h1, h2 in layer_sizes])\n",
        "        \n",
        "        self.output = nn.Linear(hidden_layers[-1], output_size)\n",
        "        \n",
        "        self.dropout = nn.Dropout(p=drop_p)\n",
        "        \n",
        "    def forward(self, x):\n",
        "        ''' Forward pass through the network, returns the output logits '''\n",
        "        \n",
        "        for each in self.hidden_layers:\n",
        "            x = F.relu(each(x))\n",
        "            x = self.dropout(x)\n",
        "        x = self.output(x)\n",
        "        \n",
        "        return F.log_softmax(x, dim=1)\n",
        "\n",
        "\n",
        "def validation(model, testloader, criterion):\n",
        "    accuracy = 0\n",
        "    test_loss = 0\n",
        "    for images, labels in testloader:\n",
        "\n",
        "        images = images.resize_(images.size()[0], 784)\n",
        "\n",
        "        output = model.forward(images)\n",
        "        test_loss += criterion(output, labels).item()\n",
        "\n",
        "        ## Calculating the accuracy \n",
        "        # Model's output is log-softmax, take exponential to get the probabilities\n",
        "        ps = torch.exp(output)\n",
        "        # Class with highest probability is our predicted class, compare with true label\n",
        "        equality = (labels.data == ps.max(1)[1])\n",
        "        # Accuracy is number of correct predictions divided by all predictions, just take the mean\n",
        "        accuracy += equality.type_as(torch.FloatTensor()).mean()\n",
        "\n",
        "    return test_loss, accuracy\n",
        "\n",
        "\n",
        "def train(model, trainloader, testloader, criterion, optimizer, epochs=5, print_every=40):\n",
        "    \n",
        "    steps = 0\n",
        "    running_loss = 0\n",
        "    for e in range(epochs):\n",
        "        # Model in training mode, dropout is on\n",
        "        model.train()\n",
        "        for images, labels in trainloader:\n",
        "            steps += 1\n",
        "            \n",
        "            # Flatten images into a 784 long vector\n",
        "            images.resize_(images.size()[0], 784)\n",
        "            \n",
        "            optimizer.zero_grad()\n",
        "            \n",
        "            output = model.forward(images)\n",
        "            loss = criterion(output, labels)\n",
        "            loss.backward()\n",
        "            optimizer.step()\n",
        "            \n",
        "            running_loss += loss.item()\n",
        "\n",
        "            if steps % print_every == 0:\n",
        "                # Model in inference mode, dropout is off\n",
        "                model.eval()\n",
        "                \n",
        "                # Turn off gradients for validation, will speed up inference\n",
        "                with torch.no_grad():\n",
        "                    test_loss, accuracy = validation(model, testloader, criterion)\n",
        "                \n",
        "                print(\"Epoch: {}/{}.. \".format(e+1, epochs),\n",
        "                      \"Training Loss: {:.3f}.. \".format(running_loss/print_every),\n",
        "                      \"Test Loss: {:.3f}.. \".format(test_loss/len(testloader)),\n",
        "                      \"Test Accuracy: {:.3f}\".format(accuracy/len(testloader)))\n",
        "                \n",
        "                running_loss = 0\n",
        "                \n",
        "                # Make sure dropout and grads are on for training\n",
        "model.train()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Classifier(\n",
              "  (fc1): Linear(in_features=784, out_features=256, bias=True)\n",
              "  (fc2): Linear(in_features=256, out_features=128, bias=True)\n",
              "  (fc3): Linear(in_features=128, out_features=64, bias=True)\n",
              "  (fc4): Linear(in_features=64, out_features=10, bias=True)\n",
              ")"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wtw8FXYSpjUa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = Network(784,10 ,[512,256,128])\n",
        "criterion = nn.NLLLoss()\n",
        "optimizer = optim.Adam(model.parameters() ,lr =0.003)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jqo4cuxeqI0h",
        "colab_type": "code",
        "outputId": "3df6eb7c-7589-4720-c778-1e360252fa86",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 816
        }
      },
      "source": [
        "train(model, trainloader, testloader, criterion, optimizer, epochs=2)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch: 1/2..  Training Loss: 0.781..  Test Loss: 0.579..  Test Accuracy: 0.794\n",
            "Epoch: 1/2..  Training Loss: 0.544..  Test Loss: 0.600..  Test Accuracy: 0.790\n",
            "Epoch: 1/2..  Training Loss: 0.572..  Test Loss: 0.528..  Test Accuracy: 0.807\n",
            "Epoch: 1/2..  Training Loss: 0.507..  Test Loss: 0.511..  Test Accuracy: 0.817\n",
            "Epoch: 1/2..  Training Loss: 0.453..  Test Loss: 0.493..  Test Accuracy: 0.825\n",
            "Epoch: 1/2..  Training Loss: 0.480..  Test Loss: 0.514..  Test Accuracy: 0.825\n",
            "Epoch: 1/2..  Training Loss: 0.496..  Test Loss: 0.514..  Test Accuracy: 0.805\n",
            "Epoch: 1/2..  Training Loss: 0.456..  Test Loss: 0.479..  Test Accuracy: 0.831\n",
            "Epoch: 1/2..  Training Loss: 0.492..  Test Loss: 0.471..  Test Accuracy: 0.825\n",
            "Epoch: 1/2..  Training Loss: 0.462..  Test Loss: 0.475..  Test Accuracy: 0.834\n",
            "Epoch: 1/2..  Training Loss: 0.425..  Test Loss: 0.497..  Test Accuracy: 0.818\n",
            "Epoch: 1/2..  Training Loss: 0.459..  Test Loss: 0.486..  Test Accuracy: 0.822\n",
            "Epoch: 1/2..  Training Loss: 0.502..  Test Loss: 0.472..  Test Accuracy: 0.833\n",
            "Epoch: 1/2..  Training Loss: 0.436..  Test Loss: 0.494..  Test Accuracy: 0.819\n",
            "Epoch: 1/2..  Training Loss: 0.444..  Test Loss: 0.477..  Test Accuracy: 0.823\n",
            "Epoch: 1/2..  Training Loss: 0.435..  Test Loss: 0.481..  Test Accuracy: 0.835\n",
            "Epoch: 1/2..  Training Loss: 0.422..  Test Loss: 0.513..  Test Accuracy: 0.812\n",
            "Epoch: 1/2..  Training Loss: 0.440..  Test Loss: 0.456..  Test Accuracy: 0.830\n",
            "Epoch: 1/2..  Training Loss: 0.431..  Test Loss: 0.441..  Test Accuracy: 0.843\n",
            "Epoch: 1/2..  Training Loss: 0.416..  Test Loss: 0.419..  Test Accuracy: 0.842\n",
            "Epoch: 1/2..  Training Loss: 0.438..  Test Loss: 0.448..  Test Accuracy: 0.839\n",
            "Epoch: 1/2..  Training Loss: 0.450..  Test Loss: 0.449..  Test Accuracy: 0.840\n",
            "Epoch: 1/2..  Training Loss: 0.427..  Test Loss: 0.460..  Test Accuracy: 0.835\n",
            "Epoch: 2/2..  Training Loss: 0.599..  Test Loss: 0.556..  Test Accuracy: 0.802\n",
            "Epoch: 2/2..  Training Loss: 0.467..  Test Loss: 0.546..  Test Accuracy: 0.812\n",
            "Epoch: 2/2..  Training Loss: 0.394..  Test Loss: 0.443..  Test Accuracy: 0.839\n",
            "Epoch: 2/2..  Training Loss: 0.409..  Test Loss: 0.455..  Test Accuracy: 0.832\n",
            "Epoch: 2/2..  Training Loss: 0.410..  Test Loss: 0.480..  Test Accuracy: 0.832\n",
            "Epoch: 2/2..  Training Loss: 0.388..  Test Loss: 0.500..  Test Accuracy: 0.819\n",
            "Epoch: 2/2..  Training Loss: 0.450..  Test Loss: 0.441..  Test Accuracy: 0.843\n",
            "Epoch: 2/2..  Training Loss: 0.408..  Test Loss: 0.439..  Test Accuracy: 0.847\n",
            "Epoch: 2/2..  Training Loss: 0.384..  Test Loss: 0.438..  Test Accuracy: 0.843\n",
            "Epoch: 2/2..  Training Loss: 0.415..  Test Loss: 0.443..  Test Accuracy: 0.843\n",
            "Epoch: 2/2..  Training Loss: 0.381..  Test Loss: 0.472..  Test Accuracy: 0.837\n",
            "Epoch: 2/2..  Training Loss: 0.405..  Test Loss: 0.459..  Test Accuracy: 0.831\n",
            "Epoch: 2/2..  Training Loss: 0.412..  Test Loss: 0.407..  Test Accuracy: 0.856\n",
            "Epoch: 2/2..  Training Loss: 0.375..  Test Loss: 0.413..  Test Accuracy: 0.848\n",
            "Epoch: 2/2..  Training Loss: 0.385..  Test Loss: 0.442..  Test Accuracy: 0.843\n",
            "Epoch: 2/2..  Training Loss: 0.404..  Test Loss: 0.405..  Test Accuracy: 0.853\n",
            "Epoch: 2/2..  Training Loss: 0.373..  Test Loss: 0.404..  Test Accuracy: 0.857\n",
            "Epoch: 2/2..  Training Loss: 0.384..  Test Loss: 0.435..  Test Accuracy: 0.847\n",
            "Epoch: 2/2..  Training Loss: 0.397..  Test Loss: 0.444..  Test Accuracy: 0.833\n",
            "Epoch: 2/2..  Training Loss: 0.398..  Test Loss: 0.445..  Test Accuracy: 0.844\n",
            "Epoch: 2/2..  Training Loss: 0.418..  Test Loss: 0.424..  Test Accuracy: 0.846\n",
            "Epoch: 2/2..  Training Loss: 0.378..  Test Loss: 0.398..  Test Accuracy: 0.859\n",
            "Epoch: 2/2..  Training Loss: 0.393..  Test Loss: 0.401..  Test Accuracy: 0.854\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vl4UaClxqeV7",
        "colab_type": "code",
        "outputId": "5075f0d3-94fb-46f4-8aa3-e63b5cfd939d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 228
        }
      },
      "source": [
        "print(\"our model:  \" , model , '\\n')\n",
        "print(model.state_dict().keys())"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "our model:   Network(\n",
            "  (hidden_layers): ModuleList(\n",
            "    (0): Linear(in_features=784, out_features=512, bias=True)\n",
            "    (1): Linear(in_features=512, out_features=256, bias=True)\n",
            "    (2): Linear(in_features=256, out_features=128, bias=True)\n",
            "  )\n",
            "  (output): Linear(in_features=128, out_features=10, bias=True)\n",
            "  (dropout): Dropout(p=0.5)\n",
            ") \n",
            "\n",
            "odict_keys(['hidden_layers.0.weight', 'hidden_layers.0.bias', 'hidden_layers.1.weight', 'hidden_layers.1.bias', 'hidden_layers.2.weight', 'hidden_layers.2.bias', 'output.weight', 'output.bias'])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "RVof5kWXqvX4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "torch.save(model.state_dict(), 'checkpoint.pth')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "reQPlNlYqzBk",
        "colab_type": "code",
        "outputId": "ced55ce7-3db4-4558-b2b0-3ecb96c27c57",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 54
        }
      },
      "source": [
        "state_dict = torch.load('checkpoint.pth')\n",
        "print(state_dict.keys())"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "odict_keys(['hidden_layers.0.weight', 'hidden_layers.0.bias', 'hidden_layers.1.weight', 'hidden_layers.1.bias', 'hidden_layers.2.weight', 'hidden_layers.2.bias', 'output.weight', 'output.bias'])\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6ThvMs7pqzvt",
        "colab_type": "code",
        "outputId": "67795848-db75-4b4f-9b6a-08204fe2601d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "model.load_state_dict(state_dict)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "IncompatibleKeys(missing_keys=[], unexpected_keys=[])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZEe-S_jY2Nu2",
        "colab_type": "code",
        "outputId": "d59967f5-c481-42c8-a260-d2e5879bd63c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "torch.cuda.torch.cuda.is_available()"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "True"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 3
        }
      ]
    }
  ]
}