nin2.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/kevinkevin556/Dlchemist/blob/main/nin2.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fXShh7z-B_2L"
      },
      "source": [
        "# **Network in Network**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rhNHsfHLyXRH"
      },
      "source": [
        "# **Section 1. Set up**"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BWl6c5NctQZm",
        "outputId": "87b5f956-8c1d-40e2-92ff-6011364af8ca"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting einops\n",
            "  Downloading einops-0.6.0-py3-none-any.whl (41 kB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m41.6/41.6 KB\u001b[0m \u001b[31m4.8 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hInstalling collected packages: einops\n",
            "Successfully installed einops-0.6.0\n"
          ]
        }
      ],
      "source": [
        "!pip install einops"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "tdLZ9y1u1eJb",
        "outputId": "b90311d7-ea46-4a5e-d429-d43c7aed0c88"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/gdrive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "-H3ltX8syaCo"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "from torch import tensor\n",
        "from tqdm import tqdm\n",
        "\n",
        "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vNy9DkTTzDnB"
      },
      "source": [
        "### *Dataset and DataLoader*\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Mo1uv42zKXCL"
      },
      "source": [
        "How to normalize CIFAR-10: [link](https://github.com/Armour/pytorch-nn-practice/blob/master/utils/meanstd.py)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "gzDqpSGUq8nK"
      },
      "outputs": [],
      "source": [
        "from torch.utils.data import Dataset, Subset\n",
        "from torchvision.datasets import CIFAR10\n",
        "from torchvision.transforms import Compose, ToTensor, ToPILImage\n",
        "from einops import rearrange, reduce\n",
        "\n",
        "\n",
        "class CIFAR10Dataset(Dataset):\n",
        "  def __init__(self, root=\"./data\", W=None,train=False, download=True, index=None):    \n",
        "    cifar10 = CIFAR10(root, train, download=download)\n",
        "    if index is None:\n",
        "      self.data, self.targets = cifar10.data, cifar10.targets\n",
        "    else:\n",
        "      self.data = cifar10.data[index]\n",
        "      self.targets = [cifar10.targets[i] for i in index]\n",
        "\n",
        "    self.normalized_data = self.global_contrast_normalize(self.data)\n",
        "    self.W = self.zca_matrix(self.normalized_data) if W is None else W\n",
        "    self.tensors = self.zca_whiten(self.normalized_data)\n",
        "\n",
        "\n",
        "  def __len__(self):\n",
        "    return len(self.data)\n",
        "\n",
        "\n",
        "  def __getitem__(self, idx):\n",
        "    image, label = self.tensors[idx], self.targets[idx]\n",
        "    return image, label\n",
        "\n",
        "\n",
        "  def zca_matrix(self, images):\n",
        "    n = len(self.targets)\n",
        "    images = rearrange(images, 'N H W C -> N (H W C)')\n",
        "    cov = torch.cov(images.T)\n",
        "    L, V = torch.linalg.eigh(cov)\n",
        "    L, V = L.real, V.real\n",
        "    W = V @ torch.diag(L**(-0.5)) @ V.T\n",
        "    return W\n",
        "\n",
        "\n",
        "  def zca_whiten(self, images):\n",
        "    ## Whiten the image\n",
        "    images = rearrange(images, 'N H W C -> N (H W C)')\n",
        "    whiten_images =  images @ self.W \n",
        "    \n",
        "    ## Tensorize the image: (dim=(c, h, w) & range=(0, 1))\n",
        "    whiten_images = rearrange(whiten_images, 'N (H W C) -> N C H W', H=32, W=32)\n",
        "    return whiten_images\n",
        "\n",
        "\n",
        "  def global_contrast_normalize(self, images, s=1, llambda=0, epsilon=1e-8):\n",
        "    images = torch.tensor(images, dtype=torch.float32)\n",
        "    mean = reduce(images, 'N C H W -> N 1 1 1', 'mean')\n",
        "    std = (reduce((images-mean)**2, 'N C H W -> N 1 1 1', 'mean') + llambda)**0.5\n",
        "    denominator = (epsilon<=std) * std + (epsilon>std) * epsilon\n",
        "    return s*(images-mean)/denominator\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BpEu7_YlqRhJ"
      },
      "source": [
        "[ZCA1](https://stats.stackexchange.com/questions/117427/what-is-the-difference-between-zca-whitening-and-pca-whitening)\n",
        "[ZCA2](https://towardsdatascience.com/pca-whitening-vs-zca-whitening-a-numpy-2d-visual-518b32033edf#:~:text=ZCA%2Dwhitening%20is%20done%20by,possible%20from%20the%20original%20data.)\n",
        "[GCN](https://blog.csdn.net/baishuo8/article/details/81943844)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 284,
          "referenced_widgets": [
            "2f524e3b30a3441993e55e28bf17de98",
            "03c6b890b97240028a9ce2168e181ce7",
            "47bb3ac9dc3b40f9bae307f793a30b5f",
            "4ac8c48416534e3bb5d5912a83ab4b9a",
            "97361734fd3f41ec9b1554f6d785c07a",
            "a3151c1fb2d947a4b3850b9c37743ded",
            "d0d833f16e06486b868a82a47469f27c",
            "a5ffcd4326bb457297195d927345971f",
            "2d88cfba6b28479da60926a145b41937",
            "71a38a45278641e191be4bfa1bcfea8c",
            "40cea8a14c1048f8bc74ddc174f1ea06"
          ]
        },
        "id": "SKsHAPLg7Po-",
        "outputId": "fe9b09f4-8fe4-4776-9640-4b5f06d29ad0"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/170498071 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "2f524e3b30a3441993e55e28bf17de98"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/cifar-10-python.tar.gz to ./data\n",
            "Files already downloaded and verified\n",
            "Files already downloaded and verified\n",
            "\n",
            "Data Size:\n",
            "* Training set   =>  40000 images\n",
            "* Validation set =>  10000 images\n",
            "* Testing set    =>  10000 images\n",
            "\n",
            "Image Shape (C, H, W):\n",
            "* Training image: torch.Size([3, 32, 32])\n",
            "* Testing image:  torch.Size([3, 32, 32])\n"
          ]
        }
      ],
      "source": [
        "# Generate indices to permute the training data and conduct train-test-split\n",
        "total_size = 50000\n",
        "valid_size = total_size // 5\n",
        "rng = np.random.default_rng(42)\n",
        "indices = rng.permutation(total_size)\n",
        "\n",
        "train_dataset = CIFAR10Dataset(train=True, index=indices[:-valid_size])\n",
        "W = train_dataset.W\n",
        "\n",
        "valid_dataset = CIFAR10Dataset(train=True, W=W, index=indices[-valid_size:])\n",
        "test_dataset = CIFAR10Dataset(train=False, W=W)\n",
        "\n",
        "print(\"\\nData Size:\")\n",
        "print(\"* Training set   => \", len(train_dataset), \"images\")\n",
        "print(\"* Validation set => \", len(valid_dataset), \"images\")\n",
        "print(\"* Testing set    => \", len(test_dataset), \"images\")\n",
        "\n",
        "print(\"\\nImage Shape (C, H, W):\")\n",
        "print(\"* Training image:\", train_dataset[0][0].shape)\n",
        "print(\"* Testing image: \", test_dataset[0][0].shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 485
        },
        "id": "Qm7BTTlj7Nc6",
        "outputId": "ea95b249-517b-4a5a-a624-ac63133b0b10"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
            "WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 1440x504 with 3 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "\n",
            "Image shape: (3, 32, 32)\n"
          ]
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "n_view = 10\n",
        "torch_dstack = lambda images: rearrange(torch.dstack(images), 'C H W -> H W C')\n",
        "\n",
        "# Sample images\n",
        "raw_images = np.hstack([train_dataset.data[i] for i in range(n_view)])\n",
        "normalized_images = np.hstack([train_dataset.normalized_data[i] for i in range(n_view)])\n",
        "images = torch_dstack([train_dataset[i][0] for i in range(n_view)])\n",
        "targets = [train_dataset[i][1] for i in range(n_view)]\n",
        "\n",
        "fig, ax = plt.subplots(3, 1, figsize=(20, 7))\n",
        "ax[0].imshow(raw_images)\n",
        "ax[0].set_title(\"Raw Images\")\n",
        "ax[0].axis('off')\n",
        "\n",
        "ax[1].imshow(normalized_images)\n",
        "ax[1].set_title(\"Global Contrast Normalized images\")\n",
        "ax[1].axis('off')\n",
        "\n",
        "ax[2].imshow(images)\n",
        "ax[2].set_title(\"ZCA Whitened Images\")\n",
        "ax[2].set_xticks(ticks=np.array(range(10))*32+16)\n",
        "ax[2].set_xticklabels(targets, {'fontsize':16})\n",
        "plt.show()\n",
        "plt.close()\n",
        "\n",
        "\n",
        "image_shape = tuple(train_dataset[0][0].shape)\n",
        "print(\"\\nImage shape:\", image_shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "CjGck5MS7xYc"
      },
      "outputs": [],
      "source": [
        "from torch.utils.data import DataLoader\n",
        "train_dataloader = DataLoader(train_dataset, batch_size=128, shuffle=True, pin_memory=True)\n",
        "valid_dataloader = DataLoader(valid_dataset, batch_size=128, shuffle=False, pin_memory=True)\n",
        "test_dataloader = DataLoader(test_dataset, batch_size=128, shuffle=False, pin_memory=True)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "q--LllAyF4jx"
      },
      "source": [
        "### *Network Architecture*"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qrLruPkYDhsB"
      },
      "source": [
        "![image.png](data:image/png;base64,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)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jxkXL0cU6Qdm"
      },
      "source": [
        "The network architecture is built based on this [prototext](https://gist.github.com/mavenlin/e56253735ef32c3c296d) from the Caffe model zoo. Note one should set `MaxPool2d(ceil_mode=True)` to get the correct dimension.\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "vTo-Ht4RF4M3",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ae75b660-4dc7-4d01-b8eb-12a0bd4165b3"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "NiN(\n",
            "  (net): Sequential(\n",
            "    (0): Sequential(\n",
            "      (0): Conv2d(3, 192, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))\n",
            "      (1): ReLU()\n",
            "      (2): Conv2d(192, 160, kernel_size=(1, 1), stride=(1, 1))\n",
            "      (3): ReLU()\n",
            "      (4): Conv2d(160, 96, kernel_size=(1, 1), stride=(1, 1))\n",
            "      (5): ReLU()\n",
            "    )\n",
            "    (1): MaxPool2d(kernel_size=3, stride=2, padding=0, dilation=1, ceil_mode=True)\n",
            "    (2): Dropout(p=0.5, inplace=False)\n",
            "    (3): Sequential(\n",
            "      (0): Conv2d(96, 192, kernel_size=(5, 5), stride=(1, 1), padding=(2, 2))\n",
            "      (1): ReLU()\n",
            "      (2): Conv2d(192, 192, kernel_size=(1, 1), stride=(1, 1))\n",
            "      (3): ReLU()\n",
            "      (4): Conv2d(192, 192, kernel_size=(1, 1), stride=(1, 1))\n",
            "      (5): ReLU()\n",
            "    )\n",
            "    (4): AvgPool2d(kernel_size=3, stride=2, padding=0)\n",
            "    (5): Dropout(p=0.5, inplace=False)\n",
            "    (6): Sequential(\n",
            "      (0): Conv2d(192, 192, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
            "      (1): ReLU()\n",
            "      (2): Conv2d(192, 192, kernel_size=(1, 1), stride=(1, 1))\n",
            "      (3): ReLU()\n",
            "      (4): Conv2d(192, 10, kernel_size=(1, 1), stride=(1, 1))\n",
            "      (5): ReLU()\n",
            "    )\n",
            "    (7): AvgPool2d(kernel_size=8, stride=1, padding=0)\n",
            "    (8): Flatten(start_dim=1, end_dim=-1)\n",
            "  )\n",
            ")\n"
          ]
        }
      ],
      "source": [
        "from torch.nn import (\n",
        "    Conv2d, ReLU, MaxPool2d, AvgPool2d, Dropout, Sequential, Flatten\n",
        ")\n",
        "\n",
        "\n",
        "def CCCP(in_channels, out_channels):\n",
        "  return Conv2d(in_channels, out_channels, kernel_size=1)\n",
        "\n",
        "def MlpConv(in_channels, hidden_channels, out_channels, kernel_size, stride=1, padding=0):\n",
        "  return nn.Sequential(\n",
        "      Conv2d(in_channels, hidden_channels[0], kernel_size, stride, padding),\n",
        "      ReLU(),\n",
        "      CCCP(hidden_channels[0], hidden_channels[1]),\n",
        "      ReLU(),\n",
        "      CCCP(hidden_channels[1], out_channels),\n",
        "      ReLU()\n",
        "  )\n",
        "\n",
        "class NiN(nn.Module):\n",
        "  def __init__(self, num_classes=10):\n",
        "    super().__init__()\n",
        "    self.net = Sequential(\n",
        "        MlpConv(3, (192, 160), 96, kernel_size=5, padding=2),\n",
        "        MaxPool2d(kernel_size=3, stride=2, ceil_mode=True),\n",
        "        Dropout(p=0.5),\n",
        "\n",
        "        MlpConv(96, (192, 192), 192, kernel_size=5, padding=2),\n",
        "        AvgPool2d(kernel_size=3, stride=2, ceil_mode=True),\n",
        "        Dropout(p=0.5),\n",
        "\n",
        "        MlpConv(192, (192, 192), num_classes, kernel_size=3, padding=1),\n",
        "        AvgPool2d(kernel_size=8, stride=1),\n",
        "        Flatten()\n",
        "    )\n",
        "\n",
        "    # Remember to initialize the weights and biases.\n",
        "    # Default initialization from pytorch doesn't work well.\n",
        "    self.net.apply(self.init_weights)\n",
        "\n",
        "  def forward(self, x):\n",
        "    return self.net(x)\n",
        "\n",
        "  @staticmethod\n",
        "  def init_weights(m):\n",
        "    if isinstance(m, Conv2d):\n",
        "        torch.nn.init.normal_(m.weight, mean=0.0, std=0.05)\n",
        "        torch.nn.init.zeros_(m.bias)\n",
        "\n",
        "print(NiN())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "Bbdt9Ki-XR1l",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "150acfa3-39b8-4106-b457-4d5a18927458"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Image shape:              [1, 3, 32, 32]\n",
            "Sequential output shape:  [1, 96, 32, 32]\n",
            "MaxPool2d output shape:   [1, 96, 16, 16]\n",
            "Dropout output shape:     [1, 96, 16, 16]\n",
            "Sequential output shape:  [1, 192, 16, 16]\n",
            "AvgPool2d output shape:   [1, 192, 8, 8]\n",
            "Dropout output shape:     [1, 192, 8, 8]\n",
            "Sequential output shape:  [1, 10, 8, 8]\n",
            "AvgPool2d output shape:   [1, 10, 1, 1]\n",
            "Flatten output shape:     [1, 10]\n"
          ]
        }
      ],
      "source": [
        "# Check forward path\n",
        "X = torch.randn(1, *image_shape)\n",
        "print(f\"{'Image shape:':25s} {list(X.shape)}\")\n",
        "for layer in NiN().net:\n",
        "    X=layer(X)\n",
        "    print(f\"{layer.__class__.__name__+' output shape:':25s} {list(X.shape)}\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "6FOS76KSblIk"
      },
      "source": [
        "### *HyperModule*"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "RaHWYi72CsFe"
      },
      "outputs": [],
      "source": [
        "import sklearn.metrics\n",
        "import pandas as pd\n",
        "import seaborn as sns\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "class HyperModule():\n",
        "    def __init__(self, model, criterion, optimizer, scheduler=None, load_path=None):\n",
        "        if load_path is None:\n",
        "            self.model, self.criterion, self.optimizer = model, criterion, optimizer\n",
        "            self.scheduler = scheduler\n",
        "            self.epoch_trained = 0\n",
        "            self.train_loss = []\n",
        "            self.valid_acc = []\n",
        "            self.test_acc = None\n",
        "        else:\n",
        "            self.load(load_path)\n",
        "\n",
        "\n",
        "    def train(self, train_dataloader, valid_dataloader, save_path=None, num_epochs=1):\n",
        "        device = torch.device('cuda')\n",
        "        self.model.to(device)\n",
        "        best_acc = 0 if len(self.valid_acc)==0 else max(self.valid_acc) \n",
        "        start_epoch = self.epoch_trained\n",
        "\n",
        "        for epoch in range(num_epochs):\n",
        "            self.model.train()\n",
        "            self.batch_loss, self.batch_acc = [], []\n",
        "            self.epoch_trained += 1\n",
        "\n",
        "            train_progress = tqdm(train_dataloader, position=0, leave=True)  ## Train\n",
        "            for images, targets in train_progress:\n",
        "                images, targets = images.to(device), targets.to(device)\n",
        "                self.update_(images, targets)\n",
        "                self.update_prog_(train_progress, start_epoch+num_epochs)\n",
        "           \n",
        "            self.update_schd_()  ## Update scheduler\n",
        "\n",
        "            self.batch_acc = self.validate(valid_dataloader) ## Validate\n",
        "            self.update_hist_()\n",
        "            print(f\"Train Loss: {np.mean(self.batch_loss):.3f}, Valid Acc:{np.mean(self.batch_acc):.3f}\")\n",
        "            \n",
        "            if np.mean(self.batch_acc) > best_acc:  ## Save the best model (if any)\n",
        "                best_acc = np.mean(self.batch_acc)\n",
        "                self.save(save_path)\n",
        "\n",
        "        self.batch_loss, self.batch_acc = [], []\n",
        "        self.test_acc = None\n",
        "        \n",
        "\n",
        "    def validate(self, dataloader):\n",
        "        device = torch.device('cuda')\n",
        "        self.model.eval()\n",
        "        batch_acc = []\n",
        "        with torch.no_grad():\n",
        "            for images, targets in dataloader:\n",
        "                images, targets = images.to(device), targets.to(device)\n",
        "                preds = self.model(images)\n",
        "                pred_labels = torch.argmax(preds, dim=1)\n",
        "                batch_acc.append((pred_labels == targets).type(torch.float32).mean().item())\n",
        "        return batch_acc\n",
        "\n",
        "\n",
        "    def test(self, dataloader, load_path=None, confusion_matrix=True, class_names=None):\n",
        "        device = torch.device('cuda')\n",
        "        np_pred_labels, np_targets = [], [] \n",
        "        if load_path is not None:\n",
        "            self.load(load_path)\n",
        "        self.model.to(device)\n",
        "        self.model.eval()\n",
        "\n",
        "        # Obtain predictions and ground truths\n",
        "        with torch.no_grad():\n",
        "            for images, targets in dataloader:\n",
        "                images, targets = images.to(device), targets.to(device)\n",
        "                pred_labels, targets = self.get_prediction_(images, targets)\n",
        "                np_pred_labels.append(pred_labels)\n",
        "                np_targets.append(targets)\n",
        "\n",
        "        np_pred_labels = np.concatenate(np_pred_labels)\n",
        "        np_targets = np.concatenate(np_targets)\n",
        "\n",
        "        if confusion_matrix:\n",
        "            conf_mat = sklearn.metrics.confusion_matrix(np_targets, np_pred_labels)\n",
        "            conf_df = self.generate_confusion_df_(conf_mat, class_names)\n",
        "            self.plot_confusion_matrix_(conf_df)\n",
        "            self.print_class_acc_(conf_df)\n",
        "        \n",
        "        ## print total accuracy\n",
        "        self.test_acc = np.mean(np_targets == np_pred_labels)\n",
        "        print(\"\\nTotal Acc:\", np.mean(np_targets == np_pred_labels))\n",
        "\n",
        "                \n",
        "    def load(self, path):\n",
        "        device = torch.device('cuda')\n",
        "        state_dict = torch.load(path)\n",
        "        \n",
        "        self.model.load_state_dict(state_dict[\"model\"])\n",
        "        self.model.to(device)\n",
        "        \n",
        "        self.optimizer.load_state_dict(state_dict[\"optimizer\"])\n",
        "        self.scheduler.load_state_dict(state_dict[\"scheduler\"])\n",
        "        self.test_acc = state_dict[\"test_acc\"]\n",
        "\n",
        "        n_train_loss = len(state_dict[\"train_loss\"])\n",
        "        n_valid_acc = len(state_dict[\"valid_acc\"])\n",
        "        epoch_trained = state_dict[\"epoch_trained\"]\n",
        "        n = min(epoch_trained, n_train_loss, n_valid_acc)\n",
        "\n",
        "        self.epoch_trained = n\n",
        "        self.train_loss = state_dict[\"train_loss\"][:n]\n",
        "        self.valid_acc = state_dict[\"valid_acc\"][:n]\n",
        "        print(\"State dict sucessfully loaded.\")\n",
        "\n",
        "\n",
        "    def save(self, path):\n",
        "        state_dict = {}\n",
        "        state_dict[\"model\"] = self.model.state_dict()\n",
        "        state_dict[\"optimizer\"] = self.optimizer.state_dict()\n",
        "        state_dict[\"scheduler\"] = self.scheduler.state_dict()\n",
        "        state_dict[\"epoch_trained\"] = self.epoch_trained\n",
        "        state_dict[\"train_loss\"] = self.train_loss\n",
        "        state_dict[\"valid_acc\"] = self.valid_acc\n",
        "        state_dict[\"test_acc\"] = self.test_acc\n",
        "        torch.save(state_dict, path)\n",
        "        print(\"State dict saved.\")\n",
        "    \n",
        "\n",
        "    def get_state_dict(self):\n",
        "        state_dict = {}\n",
        "        state_dict[\"model\"] = self.model.state_dict()\n",
        "        state_dict[\"optimizer\"] = self.optimizer.state_dict()\n",
        "        state_dict[\"scheduler\"] = self.scheduler.state_dict()\n",
        "        state_dict[\"epoch_trained\"] = self.epoch_trained\n",
        "        state_dict[\"train_loss\"] = self.train_loss\n",
        "        state_dict[\"valid_acc\"] = self.valid_acc\n",
        "        state_dict[\"test_acc\"] = self.test_acc\n",
        "        return state_dict\n",
        "\n",
        "    #--------------------------------------------------------------------------#\n",
        "\n",
        "    ### Auxilary functions in test()\n",
        "     \n",
        "    def get_prediction_(self, images, targets):\n",
        "        preds = self.model(images)\n",
        "        pred_labels = torch.argmax(preds, dim=1)\n",
        "        pred_labels = pred_labels.view(-1).detach().cpu().numpy()\n",
        "        targets = targets.view(-1).detach().cpu().numpy()\n",
        "        return pred_labels, targets\n",
        "\n",
        "    def generate_confusion_df_(self, conf_mat, class_names):\n",
        "        if class_names is not None:\n",
        "            conf_df = pd.DataFrame(conf_mat, class_names, class_names)\n",
        "        else:\n",
        "            conf_df = pd.DataFrame(conf_mat)\n",
        "        return conf_df\n",
        "        \n",
        "    def plot_confusion_matrix_(self, conf_df):\n",
        "        plt.figure(figsize = (12,8))\n",
        "        sns.heatmap(conf_df, annot=True, fmt=\"d\", cmap='Blues')\n",
        "        plt.xlabel(\"prediction\")\n",
        "        plt.ylabel(\"label (ground truth)\")\n",
        "        plt.show()\n",
        "\n",
        "    def print_class_acc_(self, conf_df):\n",
        "        for i in range(len(conf_df)):\n",
        "            total = np.sum(conf_df.iloc[i, :])\n",
        "            correct = conf_df.iloc[i, i]\n",
        "            print(f\"Acc of {conf_df.columns[i]}: {correct/total:.4f}\")\n",
        "\n",
        "    \n",
        "    ### Auxilary functions in train()\n",
        "    \n",
        "    def update_(self, images, targets):\n",
        "        preds = self.model(images)\n",
        "        loss = self.criterion(preds, targets)\n",
        "        self.optimizer.zero_grad()\n",
        "        loss.backward()\n",
        "        self.optimizer.step()\n",
        "        self.batch_loss.append(loss.detach().item())\n",
        "\n",
        "    def update_prog_(self, progress, num_epochs):\n",
        "        loss = self.batch_loss[-1]\n",
        "        epoch = self.epoch_trained\n",
        "        progress.set_description(f'Epoch [{epoch}/{num_epochs}]')\n",
        "        progress.set_postfix({'loss': loss})\n",
        "\n",
        "    def update_schd_(self):\n",
        "        if self.scheduler is not None:\n",
        "            self.scheduler.step()\n",
        "\n",
        "    def update_hist_(self):\n",
        "        # print(\"Train loss updated:\", np.mean(self.batch_loss))\n",
        "        self.train_loss.append(np.mean(self.batch_loss))\n",
        "        # print(\"Valid acc updated:\", np.mean(self.batch_acc))\n",
        "        self.valid_acc.append(np.mean(self.batch_acc))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Q0UG_Lp5fSwf"
      },
      "source": [
        "# **Section 2. Start Training**"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "uIfUjRxEfZyd"
      },
      "source": [
        "### *Optimization Configuration and First Training Stage*"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "K-1ar7-hiUwK"
      },
      "outputs": [],
      "source": [
        "BASE_LR = 0.1\n",
        "MOMENTUM = 0.9\n",
        "WEIGHT_DECAY = 1e-4\n",
        "STEP_SIZE = 1e5\n",
        "GAMMA = 0.1\n",
        "NUM_EPOCHS = 230\n",
        "PATH = \"/content/gdrive/MyDrive/Colab Checkpoints/state_dict.ckpt\""
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mLCMpxTPHVzm",
        "outputId": "063b5be7-0567-4120-b43d-42f165efcf83"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [1/230]: 100%|██████████| 313/313 [00:19<00:00, 16.30it/s, loss=2.25]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 2.243, Valid Acc:0.188\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [2/230]: 100%|██████████| 313/313 [00:19<00:00, 16.34it/s, loss=2.06]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 2.103, Valid Acc:0.270\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [3/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=1.54]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 1.770, Valid Acc:0.339\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [4/230]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=1.34]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 1.523, Valid Acc:0.456\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [5/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=1.13]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 1.302, Valid Acc:0.549\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [6/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=1.16]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 1.182, Valid Acc:0.592\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [7/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.901]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 1.053, Valid Acc:0.618\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [8/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=1.07]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.958, Valid Acc:0.676\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [9/230]: 100%|██████████| 313/313 [00:18<00:00, 16.47it/s, loss=0.676]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.884, Valid Acc:0.714\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [10/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.819]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.804, Valid Acc:0.683\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [11/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.862]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.760, Valid Acc:0.741\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [12/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.871]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.715, Valid Acc:0.756\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [13/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.773]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.659, Valid Acc:0.728\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [14/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.748]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.637, Valid Acc:0.741\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [15/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.625]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.591, Valid Acc:0.775\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [16/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.574]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.567, Valid Acc:0.785\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [17/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.402]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.534, Valid Acc:0.766\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [18/230]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.482]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.516, Valid Acc:0.782\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [19/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.566]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.485, Valid Acc:0.771\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [20/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.411]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.461, Valid Acc:0.791\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [21/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.55]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.454, Valid Acc:0.799\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [22/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.479]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.434, Valid Acc:0.786\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [23/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.417]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.418, Valid Acc:0.798\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [24/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.548]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.417, Valid Acc:0.799\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [25/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.436]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.389, Valid Acc:0.813\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [26/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.408]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.385, Valid Acc:0.814\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [27/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.499]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.373, Valid Acc:0.812\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [28/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.2]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.359, Valid Acc:0.795\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [29/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.619]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.359, Valid Acc:0.818\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [30/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.426]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.355, Valid Acc:0.819\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [31/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.415]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.338, Valid Acc:0.796\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [32/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.393]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.341, Valid Acc:0.802\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [33/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.25]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.316, Valid Acc:0.819\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [34/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.294]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.316, Valid Acc:0.819\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [35/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.2]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.308, Valid Acc:0.813\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [36/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.209]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.301, Valid Acc:0.820\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [37/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.713]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.293, Valid Acc:0.815\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [38/230]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.168]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.292, Valid Acc:0.794\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [39/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.297]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.286, Valid Acc:0.826\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [40/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.237]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.287, Valid Acc:0.822\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [41/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.406]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.293, Valid Acc:0.827\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [42/230]: 100%|██████████| 313/313 [00:18<00:00, 16.61it/s, loss=0.155]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.268, Valid Acc:0.820\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [43/230]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.262]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.270, Valid Acc:0.827\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [44/230]: 100%|██████████| 313/313 [00:18<00:00, 16.62it/s, loss=0.415]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.267, Valid Acc:0.818\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [45/230]: 100%|██████████| 313/313 [00:18<00:00, 16.63it/s, loss=0.342]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.261, Valid Acc:0.822\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [46/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.217]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.254, Valid Acc:0.829\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [47/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.211]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.254, Valid Acc:0.829\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [48/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.267]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.256, Valid Acc:0.825\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [49/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.138]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.254, Valid Acc:0.833\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [50/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.304]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.241, Valid Acc:0.819\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [51/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.332]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.240, Valid Acc:0.830\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [52/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.388]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.243, Valid Acc:0.833\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [53/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.176]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.242, Valid Acc:0.824\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [54/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.205]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.242, Valid Acc:0.830\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [55/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.273]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.237, Valid Acc:0.827\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [56/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.192]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.231, Valid Acc:0.812\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [57/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.288]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.227, Valid Acc:0.834\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [58/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.21]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.237, Valid Acc:0.823\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [59/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.299]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.227, Valid Acc:0.815\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [60/230]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.174]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.235, Valid Acc:0.822\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [61/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.326]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.219, Valid Acc:0.827\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [62/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.191]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.239, Valid Acc:0.841\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [63/230]: 100%|██████████| 313/313 [00:18<00:00, 16.60it/s, loss=0.293]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.219, Valid Acc:0.817\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [64/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.199]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.223, Valid Acc:0.830\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [65/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.266]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.213, Valid Acc:0.846\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [66/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.311]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.217, Valid Acc:0.845\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [67/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.203]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.217, Valid Acc:0.828\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [68/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.378]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.211, Valid Acc:0.832\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [69/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.17]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.213, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [70/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.29]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.234, Valid Acc:0.821\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [71/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.264]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.213, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [72/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.227]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.206, Valid Acc:0.839\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [73/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.207]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.210, Valid Acc:0.829\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [74/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.259]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.210, Valid Acc:0.834\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [75/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.247]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.195, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [76/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.213]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.207, Valid Acc:0.834\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [77/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.13]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.203, Valid Acc:0.839\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [78/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.115]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.202, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [79/230]: 100%|██████████| 313/313 [00:18<00:00, 16.60it/s, loss=0.141]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.200, Valid Acc:0.830\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [80/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.129]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.198, Valid Acc:0.832\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [81/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.236]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.199, Valid Acc:0.829\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [82/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.232]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.193, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [83/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.229]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.205, Valid Acc:0.828\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [84/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.106]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.184, Valid Acc:0.844\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [85/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.405]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.195, Valid Acc:0.828\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [86/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.218]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.197, Valid Acc:0.827\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [87/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.242]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.193, Valid Acc:0.836\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [88/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.275]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.201, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [89/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.139]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.189, Valid Acc:0.836\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [90/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.127]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.190, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [91/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.0579]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.201, Valid Acc:0.836\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [92/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.23]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.192, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [93/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.233]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.187, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [94/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.129]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.187, Valid Acc:0.845\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [95/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.118]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.191, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [96/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.316]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.188, Valid Acc:0.827\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [97/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.119]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.188, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [98/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.197]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.182, Valid Acc:0.834\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [99/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.333]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.199, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [100/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.283]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.187, Valid Acc:0.834\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [101/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.497]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.185, Valid Acc:0.831\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [102/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.253]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.189, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [103/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.166]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.192, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [104/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.305]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.184, Valid Acc:0.823\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [105/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.195]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.178, Valid Acc:0.825\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [106/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.423]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.174, Valid Acc:0.839\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [107/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.255]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.174, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [108/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.186]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.181, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [109/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.22]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.180, Valid Acc:0.845\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [110/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.154]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.175, Valid Acc:0.847\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [111/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.196]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.186, Valid Acc:0.832\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [112/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.122]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.170, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [113/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.19]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.179, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [114/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.429]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.185, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [115/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.201]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.186, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [116/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.116]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.185, Valid Acc:0.834\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [117/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.316]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.181, Valid Acc:0.854\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [118/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.18]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.175, Valid Acc:0.834\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [119/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0507]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.169, Valid Acc:0.853\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [120/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.189]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.175, Valid Acc:0.850\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [121/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.123]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.168, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [122/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.235]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.173, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [123/230]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.203]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.180, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [124/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0959]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.168, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [125/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.136]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.173, Valid Acc:0.852\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [126/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.137]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.167, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [127/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.172]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.166, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [128/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.153]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.171, Valid Acc:0.839\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [129/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.243]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.179, Valid Acc:0.847\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [130/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.248]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.183, Valid Acc:0.849\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [131/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.443]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.169, Valid Acc:0.845\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [132/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.262]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.163, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [133/230]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.107]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.170, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [134/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.194]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.159, Valid Acc:0.847\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [135/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.185]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.168, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [136/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.124]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.177, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [137/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.222]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.171, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [138/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.153]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.180, Valid Acc:0.824\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [139/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.253]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.173, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [140/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.215]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.162, Valid Acc:0.836\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [141/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.17]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.170, Valid Acc:0.850\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [142/230]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.156]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.170, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [143/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.134]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.162, Valid Acc:0.838\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [144/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.22]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.157, Valid Acc:0.850\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [145/230]: 100%|██████████| 313/313 [00:18<00:00, 16.61it/s, loss=0.246]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.163, Valid Acc:0.839\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [146/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.203]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.170, Valid Acc:0.849\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [147/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.276]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.162, Valid Acc:0.845\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [148/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.0646]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.165, Valid Acc:0.828\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [149/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.227]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.170, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [150/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.131]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.169, Valid Acc:0.856\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [151/230]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.155]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.154, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [152/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.104]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.172, Valid Acc:0.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [153/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.136]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.171, Valid Acc:0.846\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [154/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.0917]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.162, Valid Acc:0.833\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [155/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.246]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.153, Valid Acc:0.854\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [156/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.216]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.169, Valid Acc:0.832\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [157/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.342]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.164, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [158/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.135]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.166, Valid Acc:0.849\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [159/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.155]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.168, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [160/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.233]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.163, Valid Acc:0.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [161/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.127]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.163, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [162/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.134]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.164, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [163/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.159]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.156, Valid Acc:0.844\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [164/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.184]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.156, Valid Acc:0.847\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [165/230]: 100%|██████████| 313/313 [00:18<00:00, 16.61it/s, loss=0.109]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.169, Valid Acc:0.854\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [166/230]: 100%|██████████| 313/313 [00:18<00:00, 16.62it/s, loss=0.235]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.160, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [167/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.0992]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.168, Valid Acc:0.852\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [168/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.281]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.151, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [169/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.253]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.171, Valid Acc:0.839\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [170/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.135]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.160, Valid Acc:0.846\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [171/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.181]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.160, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [172/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.112]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.847\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [173/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.174]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.159, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [174/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.135]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.159, Valid Acc:0.830\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [175/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.13]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.159, Valid Acc:0.856\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [176/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.251]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.153, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [177/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.28]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.158, Valid Acc:0.849\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [178/230]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.069]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.144, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [179/230]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.14]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.167, Valid Acc:0.820\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [180/230]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.246]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.151, Valid Acc:0.852\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [181/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0897]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.154, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [182/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.128]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.157, Valid Acc:0.845\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [183/230]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.151]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.160, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [184/230]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.274]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.167, Valid Acc:0.844\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [185/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.188]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.165, Valid Acc:0.833\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [186/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.0952]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.163, Valid Acc:0.825\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [187/230]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.0383]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.157, Valid Acc:0.855\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [188/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0944]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.148, Valid Acc:0.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [189/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.25]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.154, Valid Acc:0.854\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [190/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.283]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.156, Valid Acc:0.844\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [191/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.106]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.855\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [192/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0798]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.151, Valid Acc:0.853\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [193/230]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.175]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.837\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [194/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.322]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.153, Valid Acc:0.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [195/230]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.201]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.157, Valid Acc:0.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [196/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.153]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.844\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [197/230]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.221]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.150, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [198/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.24]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.155, Valid Acc:0.844\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [199/230]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.0418]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.853\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [200/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.331]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.146, Valid Acc:0.854\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [201/230]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.169]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.146, Valid Acc:0.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [202/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.184]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.145, Valid Acc:0.851\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [203/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.203]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.159, Valid Acc:0.843\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [204/230]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.12]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.148, Valid Acc:0.853\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [205/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.307]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.155, Valid Acc:0.850\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [206/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.157]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.160, Valid Acc:0.841\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [207/230]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.223]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.149, Valid Acc:0.852\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [208/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.162]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.156, Valid Acc:0.845\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [209/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.223]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.154, Valid Acc:0.833\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [210/230]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.145]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.155, Valid Acc:0.854\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [211/230]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.15]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.857\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [212/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.176]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.158, Valid Acc:0.851\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [213/230]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.351]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.138, Valid Acc:0.847\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [214/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.2]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.165, Valid Acc:0.848\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [215/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.125]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.149, Valid Acc:0.851\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [216/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.234]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.150, Valid Acc:0.849\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [217/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.208]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.149, Valid Acc:0.857\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [218/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.206]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [219/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.15]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.151, Valid Acc:0.839\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [220/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.171]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.851\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [221/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.0472]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.853\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [222/230]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.107]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.145, Valid Acc:0.847\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [223/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.137]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.151, Valid Acc:0.851\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [224/230]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.17]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.152, Valid Acc:0.863\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [225/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.0723]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.149, Valid Acc:0.842\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [226/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.138]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.150, Valid Acc:0.824\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [227/230]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.119]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.154, Valid Acc:0.846\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [228/230]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.168]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.148, Valid Acc:0.840\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [229/230]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.0505]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.142, Valid Acc:0.850\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [230/230]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0681]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.148, Valid Acc:0.845\n"
          ]
        }
      ],
      "source": [
        "class_names = ['plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']\n",
        "\n",
        "nin = NiN(num_classes=10)\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.SGD(nin.parameters(), BASE_LR, MOMENTUM, 0, WEIGHT_DECAY)\n",
        "scheduler = torch.optim.lr_scheduler.StepLR(optimizer, STEP_SIZE, GAMMA)\n",
        "\n",
        "hm = HyperModule(\n",
        "    model = nin,\n",
        "    criterion = criterion,\n",
        "    optimizer = optimizer,\n",
        "    scheduler = scheduler\n",
        ")\n",
        "\n",
        "hm.train(train_dataloader, valid_dataloader, PATH, NUM_EPOCHS)"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "hm.test(test_dataloader, load_path=PATH, class_names=class_names)"
      ],
      "metadata": {
        "id": "JnN8fLnIcrJm",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 737
        },
        "outputId": "8ee057b8-da35-4b79-c787-b4c54640999a",
        "collapsed": true
      },
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "State dict sucessfully loaded.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x576 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Acc of plane: 0.8400\n",
            "Acc of car: 0.9020\n",
            "Acc of bird: 0.7700\n",
            "Acc of cat: 0.7400\n",
            "Acc of deer: 0.8550\n",
            "Acc of dog: 0.8040\n",
            "Acc of frog: 0.8810\n",
            "Acc of horse: 0.9100\n",
            "Acc of ship: 0.9210\n",
            "Acc of truck: 0.9360\n",
            "\n",
            "Total Acc: 0.8559\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "BASE_LR = 0.01\n",
        "MOMENTUM = 0.5\n",
        "WEIGHT_DECAY = 1e-4\n",
        "STEP_SIZE = 1e4\n",
        "GAMMA = 0.1\n",
        "NUM_EPOCHS = 100\n",
        "PATH = \"/content/gdrive/MyDrive/Colab Checkpoints/state_dict.ckpt\""
      ],
      "metadata": {
        "id": "PCiy8PTM8wua"
      },
      "execution_count": 18,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "class_names = ['plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'ship', 'truck']\n",
        "nin = NiN(num_classes=10)\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = torch.optim.SGD(nin.parameters(), BASE_LR, MOMENTUM, 0, WEIGHT_DECAY)\n",
        "scheduler = torch.optim.lr_scheduler.StepLR(optimizer, STEP_SIZE, GAMMA)\n",
        "\n",
        "hm = HyperModule(\n",
        "    model = nin,\n",
        "    criterion = criterion,\n",
        "    optimizer = optimizer,\n",
        "    scheduler = scheduler\n",
        ")\n",
        "hm.load(PATH)\n",
        "hm.optimizer = torch.optim.SGD(hm.model.parameters(), BASE_LR, MOMENTUM, 0, WEIGHT_DECAY)\n",
        "hm.scheduler = torch.optim.lr_scheduler.StepLR(hm.optimizer, STEP_SIZE, GAMMA)\n",
        "hm.train(train_dataloader, valid_dataloader, PATH, NUM_EPOCHS)"
      ],
      "metadata": {
        "id": "2CHTu9F2LjXM",
        "outputId": "cd511183-08fb-4e09-d8a8-fcdef26c5fbb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        }
      },
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "State dict sucessfully loaded.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [225/324]: 100%|██████████| 313/313 [00:18<00:00, 17.26it/s, loss=0.0566]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.076, Valid Acc:0.872\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [226/324]: 100%|██████████| 313/313 [00:18<00:00, 17.02it/s, loss=0.0195]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.059, Valid Acc:0.874\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [227/324]: 100%|██████████| 313/313 [00:18<00:00, 16.91it/s, loss=0.0125]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.050, Valid Acc:0.875\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [228/324]: 100%|██████████| 313/313 [00:18<00:00, 16.81it/s, loss=0.026]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.046, Valid Acc:0.876\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [229/324]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.0338]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.042, Valid Acc:0.877\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [230/324]: 100%|██████████| 313/313 [00:19<00:00, 16.45it/s, loss=0.0102]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.041, Valid Acc:0.879\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [231/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.0199]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.039, Valid Acc:0.879\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [232/324]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.011]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.037, Valid Acc:0.878\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [233/324]: 100%|██████████| 313/313 [00:19<00:00, 16.44it/s, loss=0.104]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.034, Valid Acc:0.878\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [234/324]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.0307]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.033, Valid Acc:0.879\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [235/324]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.0586]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.032, Valid Acc:0.879\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [236/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0119]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.029, Valid Acc:0.880\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [237/324]: 100%|██████████| 313/313 [00:18<00:00, 16.62it/s, loss=0.0212]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.029, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [238/324]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.0518]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.029, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [239/324]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.0362]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.027, Valid Acc:0.879\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [240/324]: 100%|██████████| 313/313 [00:18<00:00, 16.64it/s, loss=0.016]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.028, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [241/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.00771]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.026, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [242/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.0262]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.026, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [243/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.0674]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.026, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [244/324]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.0397]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.024, Valid Acc:0.881\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [245/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0229]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.023, Valid Acc:0.881\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [246/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.00385]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.022, Valid Acc:0.882\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [247/324]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.006]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.021, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [248/324]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.0284]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.022, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [249/324]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.0548]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.021, Valid Acc:0.882\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [250/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.00858]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.020, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [251/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0199]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.020, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [252/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.00522]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.019, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [253/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.0253]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.020, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [254/324]: 100%|██████████| 313/313 [00:18<00:00, 16.60it/s, loss=0.0611]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.019, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [255/324]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.00131]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.019, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [256/324]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.0102]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.018, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [257/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.0174]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.019, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [258/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.00438]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.019, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [259/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0641]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.017, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [260/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.016]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.018, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [261/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.00873]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.018, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [262/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.0361]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.017, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [263/324]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.00218]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.019, Valid Acc:0.880\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [264/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0248]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.016, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [265/324]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.00181]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.017, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [266/324]: 100%|██████████| 313/313 [00:18<00:00, 16.60it/s, loss=0.0778]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.016, Valid Acc:0.884\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [267/324]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.0117]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.017, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [268/324]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.0172]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.017, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [269/324]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.00882]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.017, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [270/324]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.0133]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.017, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [271/324]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.00642]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.015, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [272/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.00177]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.015, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [273/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0147]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.016, Valid Acc:0.881\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [274/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0056]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.016, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [275/324]: 100%|██████████| 313/313 [00:18<00:00, 16.59it/s, loss=0.0625]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.015, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [276/324]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.0199]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.015, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [277/324]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.00774]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.016, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [278/324]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.0287]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.015, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [279/324]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.00207]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.014, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [280/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.0514]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.014, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [281/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.00497]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [282/324]: 100%|██████████| 313/313 [00:18<00:00, 16.61it/s, loss=0.0153]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.014, Valid Acc:0.884\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [283/324]: 100%|██████████| 313/313 [00:18<00:00, 16.61it/s, loss=0.00442]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.014, Valid Acc:0.884\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [284/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0369]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.015, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [285/324]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.0123]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.015, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [286/324]: 100%|██████████| 313/313 [00:18<00:00, 16.60it/s, loss=0.0103]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.014, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [287/324]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.00058]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [288/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.0148]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [289/324]: 100%|██████████| 313/313 [00:18<00:00, 16.50it/s, loss=0.00125]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.014, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [290/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.00328]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [291/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.0493]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [292/324]: 100%|██████████| 313/313 [00:19<00:00, 16.45it/s, loss=0.0615]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [293/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.00636]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [294/324]: 100%|██████████| 313/313 [00:18<00:00, 16.54it/s, loss=0.00237]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [295/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0218]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [296/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0013]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [297/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.000971]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [298/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0135]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [299/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0489]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [300/324]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.00505]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [301/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.00256]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [302/324]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.024]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.882\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [303/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.000977]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [304/324]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.00789]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [305/324]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.00403]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.884\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [306/324]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.0197]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.013, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [307/324]: 100%|██████████| 313/313 [00:18<00:00, 16.48it/s, loss=0.00313]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.885\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [308/324]: 100%|██████████| 313/313 [00:19<00:00, 16.39it/s, loss=0.0168]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [309/324]: 100%|██████████| 313/313 [00:19<00:00, 16.45it/s, loss=0.00747]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [310/324]: 100%|██████████| 313/313 [00:19<00:00, 16.31it/s, loss=0.000401]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [311/324]: 100%|██████████| 313/313 [00:18<00:00, 16.57it/s, loss=0.00528]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [312/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.00414]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [313/324]: 100%|██████████| 313/313 [00:19<00:00, 16.47it/s, loss=0.00174]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [314/324]: 100%|██████████| 313/313 [00:18<00:00, 16.58it/s, loss=0.00346]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.885\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [315/324]: 100%|██████████| 313/313 [00:18<00:00, 16.55it/s, loss=0.0107]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.885\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [316/324]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.00245]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.885\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [317/324]: 100%|██████████| 313/313 [00:18<00:00, 16.56it/s, loss=0.00599]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [318/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.00459]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.883\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [319/324]: 100%|██████████| 313/313 [00:18<00:00, 16.52it/s, loss=0.00402]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.885\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [320/324]: 100%|██████████| 313/313 [00:18<00:00, 16.49it/s, loss=0.00594]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [321/324]: 100%|██████████| 313/313 [00:18<00:00, 16.51it/s, loss=0.00125]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [322/324]: 100%|██████████| 313/313 [00:18<00:00, 16.53it/s, loss=0.00616]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.011, Valid Acc:0.885\n",
            "State dict saved.\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [323/324]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.041]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.010, Valid Acc:0.884\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Epoch [324/324]: 100%|██████████| 313/313 [00:19<00:00, 16.46it/s, loss=0.000939]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Train Loss: 0.012, Valid Acc:0.885\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "hm.test(test_dataloader, load_path=PATH, class_names=class_names)"
      ],
      "metadata": {
        "id": "egFCY_Xta_X8",
        "outputId": "a1300519-1205-406a-91d7-ae1f6cb6ee26",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 732
        }
      },
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "State dict sucessfully loaded.\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 864x576 with 2 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Acc of plane: 0.8840\n",
            "Acc of car: 0.9450\n",
            "Acc of bird: 0.8560\n",
            "Acc of cat: 0.7390\n",
            "Acc of deer: 0.8850\n",
            "Acc of dog: 0.8160\n",
            "Acc of frog: 0.9280\n",
            "Acc of horse: 0.9170\n",
            "Acc of ship: 0.9400\n",
            "Acc of truck: 0.9190\n",
            "\n",
            "Total Acc: 0.8829\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "train_info = {\"train_loss\":hm.train_loss, \"valid_acc\":hm.valid_acc}\n",
        "\n",
        "sns.set_theme()\n",
        "fig, ax = plt.subplots(1, 1)\n",
        "sns.lineplot(ax=ax, data=pd.DataFrame(train_info))\n",
        "ax.set_xlabel(\"epochs\")\n",
        "ax.set_ylabel(\"Loss / Accuracy\")\n",
        "ax.set_title(\"Learning Curve\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 301
        },
        "id": "zbNR3Sb4lU9e",
        "outputId": "2dc98a09-6457-4814-96bb-589ac2cc231f"
      },
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "chp2qwTyPweb"
      },
      "source": [
        "Reference:\n",
        "* [Network In Network](https://arxiv.org/pdf/1312.4400.pdf)\n",
        "* [Dive into Deep Learning - Section 8.3](http://d2l.ai/chapter_convolutional-modern/nin.html)\n",
        "* [Blog Post from 星期五。見面](https://meetonfriday.com/posts/a151bfa2/)\n",
        "* [Blog Post from Math.py](https://allen108108.github.io/blog/2019/10/07/%5B%E8%AB%96%E6%96%87%5D%20Network%20In%20Network/)"
      ]
    }
  ],
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "gpuClass": "standard",
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "2f524e3b30a3441993e55e28bf17de98": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_03c6b890b97240028a9ce2168e181ce7",
              "IPY_MODEL_47bb3ac9dc3b40f9bae307f793a30b5f",
              "IPY_MODEL_4ac8c48416534e3bb5d5912a83ab4b9a"
            ],
            "layout": "IPY_MODEL_97361734fd3f41ec9b1554f6d785c07a"
          }
        },
        "03c6b890b97240028a9ce2168e181ce7": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a3151c1fb2d947a4b3850b9c37743ded",
            "placeholder": "​",
            "style": "IPY_MODEL_d0d833f16e06486b868a82a47469f27c",
            "value": "100%"
          }
        },
        "47bb3ac9dc3b40f9bae307f793a30b5f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a5ffcd4326bb457297195d927345971f",
            "max": 170498071,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_2d88cfba6b28479da60926a145b41937",
            "value": 170498071
          }
        },
        "4ac8c48416534e3bb5d5912a83ab4b9a": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_71a38a45278641e191be4bfa1bcfea8c",
            "placeholder": "​",
            "style": "IPY_MODEL_40cea8a14c1048f8bc74ddc174f1ea06",
            "value": " 170498071/170498071 [00:13&lt;00:00, 13626546.30it/s]"
          }
        },
        "97361734fd3f41ec9b1554f6d785c07a": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "a3151c1fb2d947a4b3850b9c37743ded": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d0d833f16e06486b868a82a47469f27c": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "a5ffcd4326bb457297195d927345971f": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "2d88cfba6b28479da60926a145b41937": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "71a38a45278641e191be4bfa1bcfea8c": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "40cea8a14c1048f8bc74ddc174f1ea06": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}