Update demo notebook

Author: Anton Hosgood

README.md

@@ -55,8 +55,7 @@ Standalone utility scripts.
 2. Create a **Python 3.13 or higher** virtual environment (optional but recommended):
 
    ```bash
-   python3 --version  # Ensure you have Python 3.13+ installed
-   python3 -m venv .venv
+   python3.13 -m venv .venv
    source .venv/bin/activate
    ```
 

notebooks/demo.ipynb

@@ -1,13 +1,19 @@
 {
  "cells": [
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "In this demo notebook, we train a simple model from scratch to add colour to a black-and-white image.",
+   "id": "1a2f9b3cabfd33ef"
+  },
   {
    "cell_type": "code",
    "id": "initial_id",
    "metadata": {
     "collapsed": true,
     "ExecuteTime": {
-     "end_time": "2025-05-16T17:01:44.790969Z",
-     "start_time": "2025-05-16T17:01:43.494761Z"
+     "end_time": "2025-05-20T14:46:02.736595Z",
+     "start_time": "2025-05-20T14:46:01.432375Z"
     }
    },
    "source": [
@@ -21,21 +27,108 @@
    "outputs": [],
    "execution_count": 1
   },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "Run the notebook as if it were in the project root.",
+   "id": "55935725ab2aca96"
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2025-05-20T14:46:04.058777Z",
+     "start_time": "2025-05-20T14:46:04.055123Z"
+    }
+   },
+   "cell_type": "code",
+   "source": "%cd ..",
+   "id": "f479db5d9e61248a",
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "/Users/antonhosgood/image-colorizer\n"
+     ]
+    }
+   ],
+   "execution_count": 2
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "Run the following cell to generate a sample dataset to train on.",
+   "id": "1f147f68cf7943c4"
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": "!python3 -m scripts.generate_dataset data 128 128",
+   "id": "dc10fd7b102cbc4a",
+   "outputs": [],
+   "execution_count": null
+  },
   {
    "metadata": {
     "ExecuteTime": {
-     "end_time": "2025-05-16T17:01:45.070538Z",
-     "start_time": "2025-05-16T17:01:44.847741Z"
+     "end_time": "2025-05-20T14:46:05.601372Z",
+     "start_time": "2025-05-20T14:46:05.596731Z"
     }
    },
    "cell_type": "code",
    "source": [
-    "base_dir = Path.cwd().parent\n",
+    "# To build paths relative to project root\n",
+    "base_dir = Path.cwd()\n",
     "\n",
+    "# If loading a model checkpoint\n",
     "checkpoint_path = base_dir / \"checkpoints\" / \"unet\" / \"model_epoch_10.pth\"\n",
+    "# Add your own mage to colourise\n",
+    "input_image_path = base_dir / \"jetty.jpg\"\n",
     "\n",
     "device = get_device()\n",
-    "\n",
+    "print(f\"Using device: {device}\")"
+   ],
+   "id": "cca55360ef40f85d",
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Using device: mps\n"
+     ]
+    }
+   ],
+   "execution_count": 3
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "Run the following cell to train a model from scratch (or skip it if you already have a pretrained model checkpoint).",
+   "id": "f5150439b9764d18"
+  },
+  {
+   "metadata": {},
+   "cell_type": "code",
+   "source": "!python3 -m src.train.train src/train/config.yaml",
+   "id": "14c76e35011106b2",
+   "outputs": [],
+   "execution_count": null
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "Load the trained model checkpoint.",
+   "id": "db30850aeb846bc2"
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2025-05-20T14:46:09.248453Z",
+     "start_time": "2025-05-20T14:46:09.009257Z"
+    }
+   },
+   "cell_type": "code",
+   "source": [
     "model = UNet().to(device)\n",
     "load_checkpoint(model, checkpoint_path, device)"
    ],
@@ -45,25 +138,52 @@
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "Loaded checkpoint from /Users/antonhosgood/hue-restorer/checkpoints/unet/model_epoch_10.pth\n"
+      "Loaded checkpoint from /Users/antonhosgood/image-colorizer/checkpoints/unet/model_epoch_10.pth\n"
      ]
     }
    ],
-   "execution_count": 2
+   "execution_count": 4
   },
   {
    "metadata": {
     "ExecuteTime": {
-     "end_time": "2025-05-16T17:02:02.655557Z",
-     "start_time": "2025-05-16T17:02:02.393082Z"
+     "end_time": "2025-05-20T14:46:10.126381Z",
+     "start_time": "2025-05-20T14:46:10.121870Z"
     }
    },
    "cell_type": "code",
    "source": [
-    "input_image_path = base_dir / \"jetty.jpg\"\n",
+    "num_trainable_params = sum(p.numel() for p in model.parameters() if p.requires_grad)\n",
     "\n",
-    "inference(model, device, input_image_path, resize=(256, 512))"
+    "print(f\"Number of trainable parameters: {num_trainable_params}\")"
    ],
+   "id": "ae9d1d24bd032427",
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Number of trainable parameters: 31042499\n"
+     ]
+    }
+   ],
+   "execution_count": 5
+  },
+  {
+   "metadata": {},
+   "cell_type": "markdown",
+   "source": "Run inference. If no output path is provided, a side-by-side comparison will be generated in the notebook.",
+   "id": "42bb909ebdcdca05"
+  },
+  {
+   "metadata": {
+    "ExecuteTime": {
+     "end_time": "2025-05-20T14:46:14.007293Z",
+     "start_time": "2025-05-20T14:46:13.755714Z"
+    }
+   },
+   "cell_type": "code",
+   "source": "inference(model, device, input_image_path, resize=(256, 512))",
    "id": "d29f4ab39f99f5e1",
    "outputs": [
     {
@@ -77,15 +197,7 @@
      "output_type": "display_data"
     }
    ],
-   "execution_count": 3
-  },
-  {
-   "metadata": {},
-   "cell_type": "code",
-   "outputs": [],
-   "execution_count": null,
-   "source": "",
-   "id": "b17fa37143ceba14"
+   "execution_count": 6
   }
  ],
  "metadata": {