visualize_vit_ood_cifar10.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Base settings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The autoreload extension is already loaded. To reload it, use:\n",
      "  %reload_ext autoreload\n"
     ]
    }
   ],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import time\n",
    "from tqdm import tqdm\n",
    "from pprint import pprint\n",
    "from pathlib import Path\n",
    "import pandas as pd\n",
    "from collections import OrderedDict\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torchvision.transforms as transforms\n",
    "from torch.utils.data import DataLoader\n",
    "\n",
    "device = 'cuda' if torch.cuda.is_available() else 'cpu'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# create ConfigDict from config yaml file\n",
    "import yaml\n",
    "from ml_collections import config_dict\n",
    "\n",
    "config_path = Path('configs') / 'deit_tiny-pretrained-cifar10-ood_8.yaml'\n",
    "# config_path = Path('configs') / 'custom_vit_tiny-scratch-cifar10-ood_8.yaml'\n",
    "with config_path.open('r') as f:\n",
    "    config = yaml.safe_load(f)\n",
    "    config = config_dict.ConfigDict(config)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "# frequently used variables\n",
    "model_name = config.model.name\n",
    "patch_size = config.model.patch_size\n",
    "summary = config.summary\n",
    "\n",
    "# log directories\n",
    "log_root = Path('./logs') / model_name\n",
    "checkpoint_path = log_root / 'checkpoints'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using cache found in /home/simc/.cache/torch/hub/facebookresearch_deit_main\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VisionTransformer(\n",
      "  (patch_embed): PatchEmbed(\n",
      "    (proj): Conv2d(3, 192, kernel_size=(16, 16), stride=(16, 16))\n",
      "    (norm): Identity()\n",
      "  )\n",
      "  (pos_drop): Dropout(p=0.0, inplace=False)\n",
      "  (blocks): Sequential(\n",
      "    (0): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (1): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (2): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (3): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (4): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (5): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (6): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (7): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (8): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (9): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (10): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "    (11): Block(\n",
      "      (norm1): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (attn): Attention(\n",
      "        (qkv): Linear(in_features=192, out_features=576, bias=True)\n",
      "        (attn_drop): Dropout(p=0.0, inplace=False)\n",
      "        (proj): Linear(in_features=192, out_features=192, bias=True)\n",
      "        (proj_drop): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "      (drop_path): Identity()\n",
      "      (norm2): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "      (mlp): Mlp(\n",
      "        (fc1): Linear(in_features=192, out_features=768, bias=True)\n",
      "        (act): GELU()\n",
      "        (drop1): Dropout(p=0.0, inplace=False)\n",
      "        (fc2): Linear(in_features=768, out_features=192, bias=True)\n",
      "        (drop2): Dropout(p=0.0, inplace=False)\n",
      "      )\n",
      "    )\n",
      "  )\n",
      "  (norm): LayerNorm((192,), eps=1e-06, elementwise_affine=True)\n",
      "  (pre_logits): Identity()\n",
      "  (head): Linear(in_features=192, out_features=9, bias=True)\n",
      ")\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# initialize ViT model and load pretrained weights\n",
    "from ood_with_vit.models.vit import ViT\n",
    "\n",
    "if config.model.pretrained:\n",
    "    n_class = len(config.dataset.in_distribution_class_indices)\n",
    "    model = torch.hub.load(\n",
    "        repo_or_dir=config.model.repo,\n",
    "        model=config.model.pretrained_model,\n",
    "        pretrained=False,\n",
    "    )\n",
    "    model.head = nn.Linear(model.head.in_features, n_class)\n",
    "else:\n",
    "    model = ViT(\n",
    "        image_size=config.model.img_size,\n",
    "        patch_size=config.model.patch_size,\n",
    "        num_classes=len(config.dataset.in_distribution_class_indices),\n",
    "        dim=config.model.dim_head,\n",
    "        depth=config.model.depth,\n",
    "        heads=config.model.n_heads,\n",
    "        mlp_dim=config.model.dim_mlp,\n",
    "        dropout=config.model.dropout,\n",
    "        emb_dropout=config.model.emb_dropout,\n",
    "        visualize=True,\n",
    "    )\n",
    "\n",
    "model = model.to(device=device)\n",
    "print(model)\n",
    "\n",
    "checkpoint = torch.load(checkpoint_path / f'{summary}_best.pt')\n",
    "\n",
    "state_dict = checkpoint['model_state_dict']\n",
    "trimmed_keys = []\n",
    "for key in state_dict.keys():\n",
    "    # remove prefix 'module.' for each key (in case of DataParallel)\n",
    "    trimmed_keys.append(key[7:])\n",
    "trimmed_state_dict = OrderedDict(list(zip(trimmed_keys, state_dict.values())))\n",
    "\n",
    "model.load_state_dict(trimmed_state_dict)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'airplane': 0, 'automobile': 1, 'bird': 2, 'cat': 3, 'deer': 4, 'dog': 5, 'frog': 6, 'horse': 7, 'ship': 8, 'truck': 9}\n"
     ]
    }
   ],
   "source": [
    "from ood_with_vit.datasets import OOD_CIFAR10\n",
    "\n",
    "dataset_mean, dataset_std = config.dataset.mean, config.dataset.std\n",
    "dataset_root = config.dataset.root\n",
    "img_size = config.model.img_size\n",
    "in_distribution_class_indices = config.dataset.in_distribution_class_indices\n",
    "\n",
    "transform_test = transforms.Compose([\n",
    "    transforms.Resize(img_size),\n",
    "    transforms.ToTensor(),\n",
    "    transforms.Normalize(dataset_mean, dataset_std),\n",
    "])      \n",
    "\n",
    "dataset = OOD_CIFAR10(\n",
    "    root=dataset_root, \n",
    "    in_distribution_class_indices=in_distribution_class_indices, \n",
    "    train=False, \n",
    "    download=False, \n",
    "    transform=transform_test\n",
    ")\n",
    "testloader = DataLoader(\n",
    "    dataset=dataset,\n",
    "    batch_size=config.eval.batch_size,\n",
    "    shuffle=False,\n",
    "    num_workers=8,\n",
    ")\n",
    "\n",
    "class_indices = list(range(len(dataset.classes)))\n",
    "ood_class_indices = [i for i in class_indices if i not in in_distribution_class_indices]\n",
    "in_class_indices = [i for i in class_indices if i in in_distribution_class_indices]\n",
    "print(dataset.class_to_idx)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Utility functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# set seeds\n",
    "import random\n",
    "import numpy as np\n",
    "\n",
    "def set_seed(seed):\n",
    "    random.seed(seed)\n",
    "    np.random.seed(seed)\n",
    "    torch.manual_seed(seed)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Visualize attention maps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [],
   "source": [
    "from ood_with_vit.visualizer import ViTAttentionRollout\n",
    "from ood_with_vit.utils import (\n",
    "    compute_attention_maps,\n",
    "    compute_logits,\n",
    ")\n",
    "\n",
    "def visualize_results(\n",
    "    original_img, \n",
    "    attn_weights, \n",
    "    head_fusion='max',\n",
    "    discard_ratio=0.9,\n",
    "):\n",
    "    visualizer = ViTAttentionRollout(\n",
    "        head_fusion=head_fusion,\n",
    "        discard_ratio=discard_ratio\n",
    "    )\n",
    "    mask = visualizer.rollout(attn_weights)\n",
    "\n",
    "    fig, ax = plt.subplots(figsize=(12, 6))\n",
    "    plt.subplot(1, 3, 1)\n",
    "    plt.imshow(original_img)\n",
    "\n",
    "    plt.subplot(1, 3, 2)\n",
    "    visualized_img = visualizer.get_visualized_masks(original_img, mask)\n",
    "    visualized_img = visualized_img[:, :, ::-1]\n",
    "    plt.imshow(visualized_img)\n",
    "\n",
    "    plt.subplot(1, 3, 3)\n",
    "    plt.imshow(mask)\n",
    "    cax = fig.add_axes([\n",
    "        ax.get_position().x1 + 0.01,\n",
    "        ax.get_position().y0 + ax.get_position().height * 0.2, \n",
    "        0.02,\n",
    "        ax.get_position().height * 0.6\n",
    "    ])\n",
    "    plt.colorbar(cax=cax)\n",
    "\n",
    "    plt.show()\n",
    "    \n",
    "    return mask"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Test in-distribution samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 90/90 [00:35<00:00,  2.56it/s]\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1080x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'airplane': 0.991,\n",
      " 'automobile': 0.981,\n",
      " 'bird': 0.972,\n",
      " 'cat': 0.947,\n",
      " 'deer': 0.971,\n",
      " 'dog': 0.928,\n",
      " 'frog': 0.981,\n",
      " 'horse': 0.976,\n",
      " 'truck': 0.982}\n",
      "mean acc: 0.9699\n"
     ]
    }
   ],
   "source": [
    "# calculate in-distribution data statistics\n",
    "import plotly.figure_factory as ff\n",
    "import seaborn as sn\n",
    "\n",
    "n_class = len(config.dataset.in_distribution_class_indices)\n",
    "confusion_matrix = np.zeros((n_class, n_class))\n",
    "accuracy = {}\n",
    "model.eval()\n",
    "with torch.no_grad():\n",
    "    for x, y in tqdm(testloader):\n",
    "        x, y = x.to(device), y.to(device)\n",
    "        logits = compute_logits(config, model, x)\n",
    "        _, preds = logits.max(1)\n",
    "        for label, pred in zip(y.view(-1), preds.view(-1)):\n",
    "            label, pred = int(label.item()), int(pred.item())\n",
    "            confusion_matrix[label, pred] += 1 \n",
    "\n",
    "in_dist_classes = [dataset.classes[i] for i in in_class_indices]\n",
    "df_cm = pd.DataFrame(confusion_matrix, in_dist_classes, in_dist_classes)\n",
    "plt.figure(figsize=(15, 10))\n",
    "heatmap = sn.heatmap(df_cm, annot=True, fmt='g')\n",
    "heatmap.yaxis.set_ticklabels(heatmap.yaxis.get_ticklabels(), rotation=0, ha='right', fontsize=15)\n",
    "heatmap.xaxis.set_ticklabels(heatmap.xaxis.get_ticklabels(), rotation=45, ha='right', fontsize=15)\n",
    "plt.ylabel('True label')\n",
    "plt.xlabel('Predicted label')\n",
    "plt.show()\n",
    "plt.close()\n",
    "\n",
    "accuracy = confusion_matrix.diagonal() / confusion_matrix.sum(axis=1)\n",
    "mean_acc = accuracy.mean()\n",
    "accuracy = {dataset.classes[idx]: accuracy[i] for i, idx in enumerate(in_class_indices)}\n",
    "\n",
    "pprint(accuracy)\n",
    "print(f'mean acc: {mean_acc:.4f}')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max logit: 10.042536735534668\n",
      "max class: cat\n",
      "logit: [[-1.7942 -1.001  -1.958  10.0425 -4.3302  0.0008 -1.9898 -2.8893 -0.8253]]\n",
      "attn weights: 12 torch.Size([1, 3, 197, 197])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize in-distribution class attention maps.\n",
    "plt.rcdefaults()\n",
    "np.set_printoptions(suppress=True, precision=4)\n",
    "id_class_name = 'cat' \n",
    "id_class_idx, img_idx = dataset.class_to_idx[id_class_name], 0\n",
    "original_img = dataset.img_dict[id_class_idx][img_idx]\n",
    "\n",
    "test_img = dataset.get_transformed_image(id_class_idx, img_idx)\n",
    "test_img = test_img.to(device=device)\n",
    "attention_maps = compute_attention_maps(config, model, test_img.unsqueeze(0))\n",
    "logits = compute_logits(config, model, test_img.unsqueeze(0))\n",
    "# logit, attn_weights = model(test_img.unsqueeze(0))\n",
    "max_logit, max_class = logits.max(1)\n",
    "print('max logit:', max_logit.item())\n",
    "print('max class:', dataset.classes[dataset.idx_to_id_class[max_class.item()]])\n",
    "print('logit:', logits.numpy())\n",
    "print('attn weights:', len(attention_maps), attention_maps[0].shape)\n",
    "\n",
    "visualize_results(\n",
    "    original_img, \n",
    "    attention_maps, \n",
    "    head_fusion='max',\n",
    "    discard_ratio=0.90,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Test out-of-distribution samples"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 1000/1000 [00:55<00:00, 17.95it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{'airplane': (765, 0.765),\n",
      " 'automobile': (77, 0.077),\n",
      " 'bird': (13, 0.013),\n",
      " 'cat': (4, 0.004),\n",
      " 'deer': (4, 0.004),\n",
      " 'dog': (2, 0.002),\n",
      " 'frog': (13, 0.013),\n",
      " 'horse': (3, 0.003),\n",
      " 'truck': (119, 0.119)}\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# calculate out-of-distribution data statistics\n",
    "ood_class_name = 'ship'\n",
    "ood_class_idx = dataset.class_to_idx[ood_class_name]\n",
    "counts = {dataset.classes[idx]: 0 for idx in in_class_indices}\n",
    "for i in tqdm(range(len(dataset.img_dict[ood_class_idx]))):\n",
    "    test_img = dataset.get_transformed_image(ood_class_idx, i)\n",
    "    test_img = test_img.to(device=device)\n",
    "    logits = compute_logits(config, model, test_img.unsqueeze(0))\n",
    "    _, max_class = logits.max(1)\n",
    "    max_class = dataset.classes[dataset.idx_to_id_class[max_class.item()]]\n",
    "    counts[max_class] += 1\n",
    "    \n",
    "for class_name, count in counts.items():\n",
    "    ratio = count / len(dataset.img_dict[ood_class_idx])\n",
    "    counts[class_name] = (count, ratio)\n",
    "pprint(counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "max logit: 8.236295700073242\n",
      "max class: automobile\n",
      "logit: [[ 6.0888  8.2363 -1.9129 -1.339  -3.3908 -1.3641 -1.4866 -2.2513 -0.0522]]\n",
      "attn weights: 12 torch.Size([1, 3, 197, 197])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1200x600 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# visualize out-of-distribution class attention maps.  \n",
    "ood_class_name = 'ship'\n",
    "ood_class_idx, img_idx = dataset.class_to_idx[ood_class_name], 0\n",
    "original_img = dataset.img_dict[ood_class_idx][img_idx]\n",
    "\n",
    "test_img = dataset.get_transformed_image(ood_class_idx, img_idx)\n",
    "test_img = test_img.to(device=device)\n",
    "attention_maps = compute_attention_maps(config, model, test_img.unsqueeze(0))\n",
    "logits = compute_logits(config, model, test_img.unsqueeze(0))\n",
    "max_logit, max_class = logits.max(1)\n",
    "print('max logit:', max_logit.item())\n",
    "print('max class:', dataset.classes[dataset.idx_to_id_class[max_class.item()]])\n",
    "print('logit:', logits.numpy())\n",
    "print('attn weights:', len(attention_maps), attention_maps[0].shape)\n",
    "\n",
    "mask = visualize_results(\n",
    "    original_img, \n",
    "    attention_maps, \n",
    "    head_fusion='max',\n",
    "    discard_ratio=0.90,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'numpy.ndarray'>\n",
      "mask norm: 4.0936913\n"
     ]
    }
   ],
   "source": [
    "# mask analysis\n",
    "print(type(mask))\n",
    "print('mask norm:', np.linalg.norm(mask))"
   ]
  }
 ],
 "metadata": {
  "interpreter": {
   "hash": "949777d72b0d2535278d3dc13498b2535136f6dfe0678499012e853ee9abcab1"
  },
  "kernelspec": {
   "display_name": "Python 3.9.7 64-bit",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.7"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}