project_utils.cpython-310.pyc

{"cells":[{"source":"![digitizing_team](digitizing_team.png)\n","metadata":{},"id":"307edb45-c4e3-493a-a0f1-4a9f8a588b09","cell_type":"markdown"},{"source":"DigiNsure Inc. is an innovative insurance company focused on enhancing the efficiency of processing claims and customer service interactions. Their newest initiative is digitizing all historical insurance claim documents, which includes improving the labeling of some IDs scanned from paper documents and identifying them as primary or secondary IDs.\n\nTo help them in their effort, you'll be using multi-modal learning to train an Optical Character Recognition (OCR) model. To improve the classification, the model will use **images** of the scanned documents as input and their **insurance type** (home, life, auto, health, or other). Integrating different data modalities (such as image and text) enables the model to perform better in complex scenarios, helping to capture more nuanced information. The **labels** that the model will be trained to identify are of two types: a primary and a secondary ID, for each image-insurance type pair.","metadata":{},"id":"abc661bf-f16c-4296-8715-b109dce58f4d","cell_type":"markdown"},{"source":"# ── Cell 1: Imports, data loading, and dataset visualisation ──────────────────\n\n# Standard library / third-party imports\nimport pickle\n\nimport matplotlib.pyplot as plt\nimport numpy as np\nimport torch\nimport torch.nn as nn\nfrom torch.utils.data import DataLoader\n\n# Project-specific helper (provides ProjectDataset)\nfrom project_utils import ProjectDataset\n\n# ── Load the dataset ──────────────────────────────────────────────────────────\nwith open('ocr_insurance_dataset.pkl', 'rb') as f:\n    dataset = pickle.load(f)\n\nprint(f\"Dataset loaded — {len(dataset)} samples\")\nprint(f\"Insurance types : {list(dataset.type_mapping.keys())}\")\nprint(f\"Label classes   : {list(dataset.label_mapping.keys())}\")\n\n# ── Visualisation helper ──────────────────────────────────────────────────────\ndef show_dataset_images(dataset, num_images: int = 5, seed: int = 42) -> None:\n    \"\"\"Display `num_images` random samples with their type and label.\"\"\"\n    rng     = np.random.default_rng(seed)\n    n_show  = min(num_images, len(dataset))\n    indices = rng.choice(len(dataset), size=n_show, replace=False)\n\n    fig, axes = plt.subplots(1, n_show, figsize=(4 * n_show, 4))\n    if n_show == 1:                  # subplots() returns a single Axes when n=1\n        axes = [axes]\n\n    type_keys  = list(dataset.type_mapping.keys())\n    label_keys = list(dataset.label_mapping.keys())\n    label_vals = list(dataset.label_mapping.values())\n\n    for ax, idx in zip(axes, indices):\n        img_tensor, lbl = dataset[idx]\n\n        # img_tensor[0] is the image channel; img_tensor[1] is the one-hot type\n        pixel_array  = (img_tensor[0].numpy() * 255).astype(np.uint8).reshape(64, 64)\n        ins_type     = type_keys[img_tensor[1].tolist().index(1)]\n        label_name   = label_keys[label_vals.index(lbl)]\n\n        ax.imshow(pixel_array, cmap='gray')\n        ax.axis('off')\n        ax.set_title(f\"Type: {ins_type}\\nLabel: {label_name}\", fontsize=9)\n\n    fig.suptitle(\"Sample OCR Insurance Documents\", fontsize=11, y=1.02)\n    plt.tight_layout()\n    plt.show()\n\n# ── Inspect 5 random samples ─────────────────────────────────────────────────\nshow_dataset_images(dataset, num_images=5)","metadata":{"executionCancelledAt":null,"executionTime":4437,"lastExecutedAt":1774269853792,"lastExecutedByKernel":"99a2f958-a400-4b0c-9757-a3d71751c51b","lastScheduledRunId":null,"lastSuccessfullyExecutedCode":"# ── Cell 1: Imports, data loading, and dataset visualisation ──────────────────\n\n# Standard library / third-party imports\nimport pickle\n\nimport matplotlib.pyplot as plt\nimport numpy as np\nimport torch\nimport torch.nn as nn\nfrom torch.utils.data import DataLoader\n\n# Project-specific helper (provides ProjectDataset)\nfrom project_utils import ProjectDataset\n\n# ── Load the dataset ──────────────────────────────────────────────────────────\nwith open('ocr_insurance_dataset.pkl', 'rb') as f:\n    dataset = pickle.load(f)\n\nprint(f\"Dataset loaded — {len(dataset)} samples\")\nprint(f\"Insurance types : {list(dataset.type_mapping.keys())}\")\nprint(f\"Label classes   : {list(dataset.label_mapping.keys())}\")\n\n# ── Visualisation helper ──────────────────────────────────────────────────────\ndef show_dataset_images(dataset, num_images: int = 5, seed: int = 42) -> None:\n    \"\"\"Display `num_images` random samples with their type and label.\"\"\"\n    rng     = np.random.default_rng(seed)\n    n_show  = min(num_images, len(dataset))\n    indices = rng.choice(len(dataset), size=n_show, replace=False)\n\n    fig, axes = plt.subplots(1, n_show, figsize=(4 * n_show, 4))\n    if n_show == 1:                  # subplots() returns a single Axes when n=1\n        axes = [axes]\n\n    type_keys  = list(dataset.type_mapping.keys())\n    label_keys = list(dataset.label_mapping.keys())\n    label_vals = list(dataset.label_mapping.values())\n\n    for ax, idx in zip(axes, indices):\n        img_tensor, lbl = dataset[idx]\n\n        # img_tensor[0] is the image channel; img_tensor[1] is the one-hot type\n        pixel_array  = (img_tensor[0].numpy() * 255).astype(np.uint8).reshape(64, 64)\n        ins_type     = type_keys[img_tensor[1].tolist().index(1)]\n        label_name   = label_keys[label_vals.index(lbl)]\n\n        ax.imshow(pixel_array, cmap='gray')\n        ax.axis('off')\n        ax.set_title(f\"Type: {ins_type}\\nLabel: {label_name}\", fontsize=9)\n\n    fig.suptitle(\"Sample OCR Insurance Documents\", fontsize=11, y=1.02)\n    plt.tight_layout()\n    plt.show()\n\n# ── Inspect 5 random samples ─────────────────────────────────────────────────\nshow_dataset_images(dataset, num_images=5)","outputsMetadata":{"0":{"height":80,"type":"stream"}}},"id":"8807afc3-7196-4958-8bcb-0bc68daa7a13","cell_type":"code","execution_count":1,"outputs":[{"output_type":"stream","name":"stdout","text":"Dataset loaded — 100 samples\nInsurance types : ['home', 'life', 'auto', 'health', 'other']\nLabel classes   : ['primary_id', 'secondary_id']\n"},{"output_type":"display_data","data":{"text/plain":"<Figure size 2000x400 with 5 Axes>","image/png":"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"},"metadata":{"image/png":{"width":1936,"height":413}}}]},{"source":"# Start coding here \n# ── Cell 2: Define the OCRModel ───────────────────────────────────────────────\n\nclass OCRModel(nn.Module):\n    \"\"\"\n    Multi-modal OCR classifier.\n\n    Inputs\n    ------\n    image : torch.Tensor  shape (B, 1, 64, 64)  — greyscale document image\n    ins_type : torch.Tensor  shape (B, 5)        — one-hot insurance type\n\n    Output\n    ------\n    logits : torch.Tensor  shape (B, 2)          — primary_id vs secondary_id\n    \"\"\"\n\n    def __init__(self):\n        super().__init__()\n\n        # ── Image branch (saved as image_layer per spec) ──────────────────────\n        self.image_layer = nn.Sequential(\n            # Block 1: 1 → 32 channels, spatial: 64 → 32\n            nn.Conv2d(in_channels=1, out_channels=32,\n                      kernel_size=3, padding=1),   # (B, 32, 64, 64)\n            nn.ReLU(),\n            nn.MaxPool2d(kernel_size=2),            # (B, 32, 32, 32)\n\n            # Block 2: 32 → 64 channels, spatial: 32 → 16\n            nn.Conv2d(in_channels=32, out_channels=64,\n                      kernel_size=3, padding=1),   # (B, 64, 32, 32)\n            nn.ReLU(),\n            nn.MaxPool2d(kernel_size=2),            # (B, 64, 16, 16)\n\n            # Block 3: 64 → 128 channels, spatial: 16 → 8\n            nn.Conv2d(in_channels=64, out_channels=128,\n                      kernel_size=3, padding=1),   # (B, 128, 16, 16)\n            nn.ReLU(),\n            nn.MaxPool2d(kernel_size=2),            # (B, 128,  8,  8)\n\n            nn.Flatten(),                           # (B, 128*8*8) = (B, 8192)\n        )\n\n        # ── Insurance-type branch (one-hot: 5 dims → 16) ─────────────────────\n        self.type_layer = nn.Sequential(\n            nn.Linear(5, 16),\n            nn.ReLU(),\n        )\n\n        # ── Fusion + classifier ───────────────────────────────────────────────\n        # 8192 (image) + 16 (type) → 256 → 2\n        self.classifier = nn.Sequential(\n            nn.Linear(8192 + 16, 256),\n            nn.ReLU(),\n            nn.Dropout(0.3),\n            nn.Linear(256, 2),                      # 2 classes: primary / secondary\n        )\n\n    def forward(self, image: torch.Tensor, ins_type: torch.Tensor) -> torch.Tensor:\n        img_feat  = self.image_layer(image)         # (B, 8192)\n        type_feat = self.type_layer(ins_type)       # (B, 16)\n        combined  = torch.cat([img_feat, type_feat], dim=1)  # (B, 8208)\n        return self.classifier(combined)            # (B, 2)\n\n\n# ── Quick sanity check ────────────────────────────────────────────────────────\nmodel = OCRModel()\nprint(model)\n\n# Dummy forward pass — verify shapes before training\ndummy_img  = torch.zeros(4, 1, 64, 64)   # batch of 4 greyscale 64×64 images\ndummy_type = torch.zeros(4, 5)           # batch of 4 one-hot type vectors\ndummy_out  = model(dummy_img, dummy_type)\nprint(f\"\\nOutput shape : {dummy_out.shape}\")   # expected: torch.Size([4, 2])","metadata":{"executionCancelledAt":null,"executionTime":null,"lastExecutedAt":null,"lastExecutedByKernel":null,"lastScheduledRunId":null,"lastSuccessfullyExecutedCode":null,"outputsMetadata":{"0":{"height":563,"type":"stream"}}},"id":"93887cde-b8fe-4287-b3fc-447ae7384e43","cell_type":"code","execution_count":2,"outputs":[{"output_type":"stream","name":"stdout","text":"OCRModel(\n  (image_layer): Sequential(\n    (0): Conv2d(1, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n    (1): ReLU()\n    (2): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n    (3): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n    (4): ReLU()\n    (5): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n    (6): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n    (7): ReLU()\n    (8): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n    (9): Flatten(start_dim=1, end_dim=-1)\n  )\n  (type_layer): Sequential(\n    (0): Linear(in_features=5, out_features=16, bias=True)\n    (1): ReLU()\n  )\n  (classifier): Sequential(\n    (0): Linear(in_features=8208, out_features=256, bias=True)\n    (1): ReLU()\n    (2): Dropout(p=0.3, inplace=False)\n    (3): Linear(in_features=256, out_features=2, bias=True)\n  )\n)\n\nOutput shape : torch.Size([4, 2])\n"}]}],"metadata":{"colab":{"name":"Welcome to DataCamp Workspaces.ipynb","provenance":[]},"kernelspec":{"display_name":"Python 3 (ipykernel)","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":"3.10.12"}},"nbformat":4,"nbformat_minor":5}