Merge pull request #115 from HeidiSteen/main

RAG quickstart updates

Author: HeidiSteen

Quickstart-RAG/Quickstart-rag.ipynb

@@ -25,13 +25,31 @@
    "source": [
     "## Prerequisites\n",
     "\n",
-    "- Azure AI Search, Basic tier or higher is recommended. Enable semantic ranking and role-based access control.\n",
+    "- [Azure OpenAI](https://learn.microsoft.com/azure/ai-services/openai/how-to/create-resource)\n",
     "\n",
-    "- Azure OpenAI, in the same region as Azure AI Search. Configure Azure OpenAI to use a system managed identity.\n",
+    "  - Configure Azure OpenAI to use a system managed identity.\n",
+    "  - Deploy a chat model (GPT-3.5-Turbo, GPT-4, or equivalent LLM).\n",
     "\n",
-    "- Deployment GPT-3.5-Turbo, GPT-4, or equivalent LLM.\n",
+    "- [Azure AI Search](https://learn.microsoft.com/azure/search/search-create-service-portal)\n",
     "\n",
-    "- Create the hotels-sample-index using [this quickstart](https://learn.microsoft.com/azure/search/search-get-started-portal), and then modify it to use this semantic configuration:\n",
+    "  - Basic tier or higher is recommended.\n",
+    "  - Choose the same region as Azure OpenAI.\n",
+    "  - Enable semantic ranking.\n",
+    "  - Enable role-based access control.\n",
+    "\n",
+    "## Configure access\n",
+    "\n",
+    "This quickstart assumes authentication and authorization using Microsoft Entra ID and role assignments. It also assumes that you run this code from your local device.\n",
+    "\n",
+    "1. To create and load the sample index on Azure AI Search, you must have role assignments for: **Search Index Data Reader**, **Search Index Data Contributor**, **Search Service Contributor**.\n",
+    "\n",
+    "1. To send the query and results from the search engine to Azure OpenAI, the search system identity must have **Cognitive Services OpenAI User** permissions on Azure OpenAI.\n",
+    "\n",
+    "## Create the sample index\n",
+    "\n",
+    "This quickstart assumes the hotels-sample-index, which you can create in minutes using [this quickstart](https://learn.microsoft.com/azure/search/search-get-started-portal).\n",
+    "\n",
+    "Once the index exists, modify it in the Azure portal to use this semantic configuration:\n",
     "  \n",
     "    ```json\n",
     "    \"semantic\": {\n",
@@ -49,38 +67,47 @@
     "       }]},\n",
     "    ```\n",
     "\n",
-    "- On Azure AI Search, create a role assignment for the Azure OpenAI managed identity. Required roles: **Search Index Data Reader**, **Search Service Contributor**.\n",
+    "Now that you have your Azure resources, an index, and model in place, you can run the script to chat with the index."
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "## Run the code\n",
+    "\n",
+    "1. Create a virtual environment. In Visual Studio Code, press Ctrl-shift-P to open the command palette, search for \"Python: Create Environment\", and then select `Venv` to create a virtual environment in the current workspace.\n",
+    "\n",
+    "1. Select Quickstart-RAG\\requirements.txt for the dependencies.\n",
     "\n",
-    "- On Azure OpenAI, create a role assigment for the user running the code: Required role: **Cognitive Services OpenAI User**."
+    "It takes several minutes to create the environment. When the environment is ready, continue to the next step."
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [],
    "source": [
     "# Package install for quickstart\n",
-    "! pip install azure-search-documents==11.6.0b4 --quiet\n",
-    "! pip install azure-identity==1.16.0 --quiet\n",
-    "! pip install openai --quiet"
+    "! pip install -r requirements.txt --quiet"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [],
    "source": [
-    "# Set endpoints and deployment model\n",
-    "AZURE_SEARCH_SERVICE: str = \"PUT YOUR SEARCH SERVICE ENDPOINT HERE\"\n",
-    "AZURE_OPENAI_ACCOUNT: str = \"PUT YOUR AZURE OPENAI ENDPOINT HERE\"\n",
-    "AZURE_DEPLOYMENT_MODEL: str = \"gpt-35-turbo\""
+    "# Set endpoints and deployment model (provide the name of the deployment)\n",
+    " AZURE_SEARCH_SERVICE: str = \"PUT YOUR SEARCH SERVICE ENDPOINT HERE\"\n",
+    " AZURE_OPENAI_ACCOUNT: str = \"PUT YOUR AZURE OPENAI ENDPOINT HERE\"\n",
+    " AZURE_DEPLOYMENT_MODEL: str = \"gpt-35-turbo\""
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -92,7 +119,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 16,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -185,9 +212,21 @@
   },
   {
    "cell_type": "code",
-   "execution_count": null,
+   "execution_count": 19,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Based on your preferences, I recommend the following hotels: \n",
+      "\n",
+      "- Ocean Air Motel: This hotel is oceanfront with beach access, two pools, and a private balcony with ocean views. \n",
+      "- Marquis Plaza & Suites: This hotel has a view of the ocean and amenities like free Wi-Fi, a full kitchen, and a free breakfast buffet. \n",
+      "- Trails End Motel: Though not directly on the ocean, this hotel does have a view and is only 8 miles from downtown. Amenities include an on-site bar/restaurant, free hot breakfast buffet, and free Wi-Fi.\n"
+     ]
+    }
+   ],
    "source": [
     "# Instantiate the chat thread and run the conversation\n",
     "import azure.identity.aio\n",
@@ -201,12 +240,19 @@
     "        query=\"Can you recommend a few hotels near the ocean with beach access and good views\",\n",
     "        search_type=SearchType(search_type),\n",
     "        use_semantic_reranker=use_semantic_reranker,\n",
-    "        sources_to_include=sources_to_include,\n",
-    "        k=k)\n",
+    "        sources_to_include=sources_to_include\n",
+    "        )\n",
     "    await chat_thread.get_openai_response(openai_client=openai_client, model=chat_deployment)\n",
     "\n",
     "print(chat_thread.get_last_message()[\"content\"])"
    ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "If you get an authorization error message, wait a few minutes and try again. It can take several minutes for role assignments to become operational."
+   ]
   }
  ],
  "metadata": {

Quickstart-RAG/requirements.txt

@@ -0,0 +1,4 @@
+azure-search-documents==11.6.0b4
+azure-identity==1.16.0 
+openai
+aiohttp