For blog - Create multimodal_mountain_peak_search.ipynb

notebooks/multimodal_mountain_peak_search/multimodal_mountain_peak_search.ipynb

@@ -0,0 +1,218 @@
+{
+  "nbformat": 4,
+  "nbformat_minor": 5,
+  "cells": [
+    {
+      "cell_type": "markdown",
+      "metadata": {
+        "id": "ppy7so3HYiOK"
+      },
+      "source": [
+        "# Multimodal Mountain Peak Search — Colab Notebook\n",
+        "\n",
+        "In this notebook we do the following:\n",
+        "- Install dependencies\n",
+        "- Connect to **Elasticsearch**\n",
+        "- Create indices\n",
+        "- Index a small **peaks catalog** (text+image blended vectors)\n",
+        "- Index a few **photos**\n",
+        "- Run **text → image** search and **identify-from-photo** search\n"
+      ],
+      "id": "ppy7so3HYiOK"
+    },
+    {
+      "cell_type": "code",
+      "metadata": {
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "id": "o0tHMSj6YiOM",
+        "outputId": "f58fb49f-1d7f-49a3-cb26-07f63534497d",
+        "cellView": "form"
+      },
+      "execution_count": 23,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Elasticsearch url(https://localhost:9200): \n",
+            "Elasticsearch base64 api keyw\n",
+            "ES_URL     : https://localhost:9200\n",
+            "API_KEY_B64: set\n"
+          ]
+        }
+      ],
+      "source": [
+        "#@title Configure Elasticsearch (API key recommended)\n",
+        "import os, base64\n",
+        "\n",
+        "# Prompt for Elasticsearch URL:\n",
+        "ES_URL = input(\"Elasticsearch url(https://localhost:9200): \") or \"https://localhost:9200\"\n",
+        "\n",
+        "\n",
+        "#Prompt Elasticsearch base64 api key:\n",
+        "ES_API_KEY_B64 = input(\"Elasticsearch base64 api key\")\n",
+        "\n",
+        "\n",
+        "print(\"ES_URL     :\", ES_URL)\n",
+        "print(\"API_KEY_B64:\", \"set\" if ES_API_KEY_B64 else \"MISSING\")\n",
+        "\n",
+        "# Propagate for scripts\n",
+        "os.environ[\"ES_URL\"] = ES_URL\n",
+        "os.environ[\"ES_API_KEY_B64\"] = ES_API_KEY_B64\n",
+        "\n",
+        "# Optional model override\n",
+        "# os.environ[\"SIGLIP_MODEL_ID\"] = \"google/siglip-so400m-patch14-384\"\n"
+      ],
+      "id": "o0tHMSj6YiOM"
+    },
+    {
+      "cell_type": "code",
+      "metadata": {
+        "id": "za6W0opPYiOM",
+        "colab": {
+          "base_uri": "https://localhost:8080/"
+        },
+        "outputId": "bdbdac01-1579-4b60-b2b1-f753deaf9548"
+      },
+      "execution_count": 2,
+      "outputs": [
+        {
+          "output_type": "stream",
+          "name": "stdout",
+          "text": [
+            "Cloning: https://github.com/navneet83/multimodal-mountain-peak-search\n",
+            "Cloned and cwd set to /content/multimodal-mountain-peak-search\n"
+          ]
+        }
+      ],
+      "source": [
+        "#@title Clone Github repo\n",
+        "REPO_URL = \"https://github.com/navneet83/multimodal-mountain-peak-search\"\n",
+        "TARGET_DIR = \"/content/multimodal-mountain-peak-search\"\n",
+        "import os, shutil, subprocess, sys\n",
+        "\n",
+        "if os.path.exists(TARGET_DIR):\n",
+        "    shutil.rmtree(TARGET_DIR)\n",
+        "\n",
+        "print(\"Cloning:\", REPO_URL)\n",
+        "rc = subprocess.call([\"git\",\"clone\",\"--depth\",\"1\", REPO_URL, TARGET_DIR])\n",
+        "if rc != 0:\n",
+        "    raise SystemExit(\"❌ Clone failed. Check the repo URL or network.\")\n",
+        "\n",
+        "os.chdir(TARGET_DIR)\n",
+        "sys.path.insert(0, os.path.join(TARGET_DIR, \"src\"))  # import ai_mpi.embeddings\n",
+        "print(\"Cloned and cwd set to\", TARGET_DIR)"
+      ],
+      "id": "za6W0opPYiOM"
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "#@title Install dependencies\n",
+        "!pip -q install --upgrade pip\n",
+        "!pip install -r requirements.txt\n",
+        "print(\"Installed\")"
+      ],
+      "metadata": {
+        "id": "JEMGE9kgrR5v"
+      },
+      "id": "JEMGE9kgrR5v",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "metadata": {
+        "id": "6T5tDn7ZYiOO"
+      },
+      "execution_count": null,
+      "outputs": [],
+      "source": [
+        "#@title Create indices in Elasticsearch\n",
+        "!python scripts/create_indices.py --recreate || python scripts/create_indices.py\n",
+        "print(\"Indices ready\")"
+      ],
+      "id": "6T5tDn7ZYiOO"
+    },
+    {
+      "cell_type": "code",
+      "metadata": {
+        "id": "tk6MFsDyYiOP"
+      },
+      "execution_count": null,
+      "outputs": [],
+      "source": [
+        "#@title Index peaks (blended text + reference images)\n",
+        "!python scripts/embed_and_index_photos.py --index-peaks --peaks-yaml data/peaks.yaml --peaks-images-root data/peaks --blend-alpha-text 0.55 --blend-max-images 3\n",
+        "print(\"Peaks indexed\")"
+      ],
+      "id": "tk6MFsDyYiOP"
+    },
+    {
+      "cell_type": "code",
+      "metadata": {
+        "id": "ualslgxsYiOP"
+      },
+      "execution_count": null,
+      "outputs": [],
+      "source": [
+        "#@title Index your photos\n",
+        "!python scripts/embed_and_index_photos.py --index-photos --images data/images --topk-predicted 5\n",
+        "print(\"✅ Photos indexed\")"
+      ],
+      "id": "ualslgxsYiOP"
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "#@title Text → image search (type a peak name)\n",
+        "query = \"Pumori\"  #@param [\"Ama Dablam\", \"Pumori\", \"Mount Everest\"] {allow-input: true}\n",
+        "k = 12  #@param {type:\"slider\", min:6, max:30, step:2}\n",
+        "num_candidates = 4000  #@param {type:\"slider\", min:1000, max:6000, step:1000}\n",
+        "\n",
+        "!python scripts/query_by_peak_name.py --peak query --k {k} --num-candidates {num_candidates}"
+      ],
+      "metadata": {
+        "cellView": "form",
+        "id": "VjpCGK9Csz2F"
+      },
+      "id": "VjpCGK9Csz2F",
+      "execution_count": null,
+      "outputs": []
+    },
+    {
+      "cell_type": "code",
+      "source": [
+        "#@title Identify from photo → similar photos (upload OR reuse a repo image)\n",
+        "from google.colab import files\n",
+        "uploaded = files.upload()\n",
+        "\n",
+        "for fn in uploaded.keys():\n",
+        "  !python scripts/identify_from_picture_find_similar_peaks.py --image \"{fn}\" --neighbors 30\n"
+      ],
+      "metadata": {
+        "id": "mLirCeRJrkDV"
+      },
+      "id": "mLirCeRJrkDV",
+      "execution_count": null,
+      "outputs": []
+    }
+  ],
+  "metadata": {
+    "colab": {
+      "name": "02_quickstart_colab_navneet_v2.ipynb",
+      "provenance": []
+    },
+    "kernelspec": {
+      "display_name": "Python 3",
+      "language": "python",
+      "name": "python3"
+    },
+    "language_info": {
+      "name": "python",
+      "version": "3.x"
+    }
+  }
+}