{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/duoan/TorchCode/blob/master/templates/22_conv2d.ipynb)\n",
    "\n",
    "# 🟠 Medium: 2D Convolution\n",
    "\n",
    "Implement **2D convolution** from scratch.\n",
    "\n",
    "### Signature\n",
    "```python\n",
    "def my_conv2d(x, weight, bias=None, stride=1, padding=0):\n",
    "    # x: (B, C_in, H, W), weight: (C_out, C_in, kH, kW)\n",
    "    # Returns: (B, C_out, H_out, W_out)\n",
    "```\n",
    "\n",
    "### Rules\n",
    "- Do NOT use `F.conv2d` or `nn.Conv2d`\n",
    "- Support `stride` and `padding` parameters\n",
    "- `F.pad` for zero-padding is allowed"
   ],
   "outputs": []
  },
  {
   "cell_type": "code",
   "metadata": {},
   "source": [
    "# Install torch-judge in Colab (no-op in JupyterLab/Docker)\n",
    "try:\n",
    "    import google.colab\n",
    "    get_ipython().run_line_magic('pip', 'install -q torch-judge')\n",
    "except ImportError:\n",
    "    pass\n"
   ],
   "outputs": [],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "import torch.nn.functional as F"
   ],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ✏️ YOUR IMPLEMENTATION HERE\n",
    "\n",
    "def my_conv2d(x, weight, bias=None, stride=1, padding=0):\n",
    "    pass  # extract patches, apply kernel, handle stride/padding"
   ],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 🧪 Debug\n",
    "x = torch.randn(1, 3, 8, 8)\n",
    "w = torch.randn(16, 3, 3, 3)\n",
    "print('Output:', my_conv2d(x, w).shape)\n",
    "print('Match:', torch.allclose(my_conv2d(x, w), F.conv2d(x, w), atol=1e-4))"
   ],
   "execution_count": null
  },
  {
   "cell_type": "code",
   "metadata": {},
   "outputs": [],
   "source": [
    "# ✅ SUBMIT\n",
    "from torch_judge import check\n",
    "check('conv2d')"
   ],
   "execution_count": null
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "name": "python",
   "version": "3.11.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}