Merge pull request #68 from Azure/jan24

Significant update - latest packages and libraries, multiple changes

.env.example

@@ -1,12 +1,10 @@
-OPENAI_API_TYPE = "Set this to "azure" for API key authentication or "azure_ad" for Azure AD authentication>"
-OPENAI_API_KEY = "<YOUR AZURE OPENAI API KEY - If using Azure AD auth, this can be left empty>"
-OPENAI_API_BASE = "<YOUR AZURE OPENAI ENDPOINT>"
-OPENAI_API_VERSION = "2023-05-15"
-OPENAI_COMPLETION_MODEL = "<YOUR OPENAI COMPLETIONS MODEL NAME - e.g. gpt-35-turbo>"
-AZURE_TENANT_ID = "<AZURE AD TENANT ID - Only used if you are using Azure AD to authentication>"
+OPENAI_API_VERSION = "2023-09-01-preview"
+AZURE_OPENAI_API_KEY = "<YOUR AZURE OPENAI API KEY - If using Azure AD auth, this can be left empty>"
+AZURE_OPENAI_ENDPOINT = "<YOUR AZURE OPENAI ENDPOINT>"
+AZURE_OPENAI_COMPLETION_MODEL = "<YOUR OPENAI COMPLETIONS MODEL NAME - e.g. gpt-35-turbo>"
 AZURE_OPENAI_COMPLETION_DEPLOYMENT_NAME = "<YOUR AZURE OPENAI COMPLETIONS DEPLOYMENT NAME - e.g. gpt-35-turbo>"
 AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME = "<YOUR AZURE OPENAI EMBEDDINGS DEPLOYMENT NAME - e.g. text-embedding-ada-002>"
-AZURE_COGNITIVE_SEARCH_SERVICE_NAME = "<YOUR AZURE COGNITIVE SEARCH SERVICE NAME - e.g. cognitive-search-service>"
-AZURE_COGNITIVE_SEARCH_ENDPOINT_NAME = "<YOUR AZURE COGNITIVE SEARCH ENDPOINT NAME - e.g. https://cognitive-search-service.search.windows.net"
-AZURE_COGNITIVE_SEARCH_INDEX_NAME = "<YOUR AZURE COGNITIVE SEARCH INDEX NAME - e.g. cognitive-search-index>"
-AZURE_COGNITIVE_SEARCH_API_KEY = "<YOUR AZURE COGNITIVE SEARCH ADMIN API KEY - e.g. cognitive-search-admin-api-key>"
+AZURE_AI_SEARCH_SERVICE_NAME = "<YOUR AZURE AI SEARCH SERVICE NAME - e.g. ai-vectorstore-xyz>"
+AZURE_AI_SEARCH_ENDPOINT = "<YOUR AZURE AI SEARCH ENDPOINT NAME - e.g. https://ai-vectorstore-xyz.search.windows.net"
+AZURE_AI_SEARCH_INDEX_NAME = "<YOUR AZURE AI SEARCH INDEX NAME - e.g. ai-search-index>"
+AZURE_AI_SEARCH_API_KEY = "<YOUR AZURE AI SEARCH ADMIN API KEY - get this value from the Azure portal>"

README.md

@@ -4,12 +4,12 @@ This repository introduces and helps organizations get started with building AI
 
 ## Workshop Agenda
 
-The objective of this workshop is to practice realistic AI orchestration scenarios towards learning how to build intelligent apps.
+The objective of this workshop is to practice realistic AI orchestration scenarios and to learn how to build intelligent apps.
 At the end of the workshop you will: 
 * Know how to use prompt engineering techniques for effective generative AI responses on OpenAI
 * Understand the implications of the usage of tokens and embeddings when interacting with an LLM
 * Have experience in leveraging AI orchestrators like Langchain/ Semantic Kernel with Azure OpenAI
-* Have evaluated different Vectorstores like Qdrant/ Azure Cognitive Search to enhance LLM responses with your data and context
+* Have evaluated different vector stores like Qdrant or Azure AI Search to enhance LLM responses with your data and context
 * Know how to turn a business scenario with data, context and user input into an intelligent application on Azure
 
 ### 🌅 Morning (9:00 – 12:15)

labs/00-setup/README.md

@@ -4,7 +4,7 @@ Here's a detailed walkthrough of what you need to get started with the labs.
 
 ## OpenAI Model Deployments
 
-During these labs, we'll make use of a *Completions* model and an *Embeddings* model, so you'll need access to at least one of each of these. Grant the participants access to Azure OpenAI Service service be assigning the `Cognitive Service OpenAI user`. If the participant is a `Cognitive Service OpenAI contributor`, they can create the following deployments themselves.
+During these labs, we'll make use of a *Completions* model and an *Embeddings* model, so you'll need access to at least one of each of these. Grant the participants access to Azure OpenAI Service service by assigning the `Cognitive Service OpenAI user`. If the participant is a `Cognitive Service OpenAI contributor`, they can create the following deployments themselves.
 
 **NOTE:** Please take extra care to make sure the correct version of each model is deployed.
 

labs/02-integrating-ai/00-PythonModules/pythonmodules.ipynb

@@ -6,7 +6,7 @@
    "source": [
     "# Python Modules\n",
     "\n",
-    "A handful of Python modules are necessary to use the labs in this repo. The module names are listed in the `requirements.txt` file which can be found in the `labs` folder.\n",
+    "A handful of Python modules are necessary to use the labs in this repo. The module names are listed in the `requirements.txt` file which can be found in the root of this repository.\n",
     "\n",
     "Run the below to install the required packages.\n",
     "\n",
@@ -55,7 +55,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.5"
+   "version": "3.11.7"
   },
   "orig_nbformat": 4
  },

labs/02-integrating-ai/01-AzureOpenAIAPI/azureopenaiapi.ipynb

@@ -31,87 +31,13 @@
     "\n",
     "# Load environment variables\n",
     "if load_dotenv():\n",
-    "    print(\"Found OpenAI API Base Endpoint: \" + os.getenv(\"OPENAI_API_BASE\"))\n",
+    "    print(\"Found Azure OpenAI API Base Endpoint: \" + os.getenv(\"AZURE_OPENAI_ENDPOINT\"))\n",
     "else: \n",
-    "    print(\"OpenAI API Base Endpoint not found. Have you configured the .env file?\")\n",
+    "    print(\"Azure OpenAI API Base Endpoint not found. Have you configured the .env file?\")\n",
     "    \n",
-    "API_KEY = os.getenv(\"OPENAI_API_KEY\")\n",
-    "# This version of the API is needed to properly retrieve the list of model deployments.\n",
-    "API_VERSION = \"2023-03-15-preview\"\n",
-    "RESOURCE_ENDPOINT = os.getenv(\"OPENAI_API_BASE\")"
-   ]
-  },
-  {
-   "attachments": {},
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Next, we'll set the endpoint that we're going to call. This endpoint will return a list of available model deployments."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "url = RESOURCE_ENDPOINT + \"/openai/deployments?api-version=\" + API_VERSION\n",
-    "\n",
-    "print (url)"
-   ]
-  },
-  {
-   "attachments": {},
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Above you will see the full URL that we are going to be calling."
-   ]
-  },
-  {
-   "attachments": {},
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "## Get the list of available Model Deployments\n",
-    "\n",
-    "Now let's call that Azure OpenAI endpoint and see what it returns. We pass the API key in the HTTP header."
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "r = requests.get(url, headers={\"api-key\": API_KEY})\n",
-    "\n",
-    "print(r.text)"
-   ]
-  },
-  {
-   "attachments": {},
-   "cell_type": "markdown",
-   "metadata": {},
-   "source": [
-    "Above, there should now be a JSON formatted list of model deployments. The output will contain one or more sections similar to the following\n",
-    "\n",
-    "```\n",
-    "    {\n",
-    "      \"scale_settings\": {\n",
-    "        \"scale_type\": \"standard\"\n",
-    "      },\n",
-    "      \"model\": \"gpt-35-turbo\",\n",
-    "      \"owner\": \"organization-owner\",\n",
-    "      \"id\": \"gpt35turbo\",\n",
-    "      \"status\": \"succeeded\",\n",
-    "      \"created_at\": 1684150536,\n",
-    "      \"updated_at\": 1684150536,\n",
-    "      \"object\": \"deployment\"\n",
-    "    }\n",
-    "```\n",
-    "\n",
-    "In the output, you'll see the **Model Name** as the `model` value and the **Deployment Name** as the `id` value. We'll be using an `id` value for the next section."
+    "API_KEY = os.getenv(\"AZURE_OPENAI_API_KEY\")\n",
+    "API_VERSION = os.getenv(\"OPENAI_API_VERSION\")\n",
+    "RESOURCE_ENDPOINT = os.getenv(\"AZURE_OPENAI_ENDPOINT\")"
    ]
   },
   {
@@ -121,9 +47,7 @@
    "source": [
     "## Send a prompt to Azure OpenAI using the API\n",
     "\n",
-    "Next, let's call the Azure OpenAI API with a prompt. To do this, we'll need the `id` of our completion model deployment.\n",
-    "\n",
-    "We've already setup this value in the `.env` file, but you should see a matching value in the list that was output above."
+    "Now let's call the Azure OpenAI API with a prompt. To do this, we'll need the `id` of the Azure OpenAI deployment that contains our completion model. This should already be setup in the `.env` file."
    ]
   },
   {
@@ -132,8 +56,6 @@
    "metadata": {},
    "outputs": [],
    "source": [
-    "# Grab API Version in the .env file.\n",
-    "API_VERSION = os.getenv(\"OPENAI_API_VERSION\")\n",
     "DEPLOYMENT_ID = os.getenv(\"AZURE_OPENAI_COMPLETION_DEPLOYMENT_NAME\")"
    ]
   },
@@ -142,7 +64,7 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "As before, we'll construct a URL to call. This time, we'll also be creating a payload."
+    "We'll construct a URL to call so that you can see what it looks like."
    ]
   },
   {
@@ -161,9 +83,9 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "Again, you will see the full URL that we are going to call. This URL will use the specified deployment to call the **chat completions** API.\n",
+    "You will see the full URL that we are going to call. This URL will use the specified deployment to call the **chat completions** API.\n",
     "\n",
-    "Next, we will call the Azure OpenAI API using the URL above. Just like last time, we pass the API key in the HTTP header. We also send a JSON formatted body as part of the request which contains the *prompt* that we want to use to get a response from the OpenAI model. In this case, our prompt is \"Once upon a time\", which should cause the model to complete the prompt by generating a story."
+    "Next, we will call the Azure OpenAI API using the URL above. We pass the API key in the HTTP header. We also send a JSON formatted body as part of the request which contains the *prompt* that we want to use to get a response from the OpenAI model. In this case, our prompt is \"Once upon a time\", which should cause the model to complete the prompt by generating a story."
    ]
   },
   {
@@ -182,14 +104,20 @@
    "cell_type": "markdown",
    "metadata": {},
    "source": [
-    "As before, the result of the API call will be JSON data similar to the below example.\n",
+    "The result of the API call will be JSON data similar to the below example. Note that the response has been edited to make it easier to read.\n",
     "\n",
     "```\n",
     "{\n",
     "  \"id\": \"chatcmpl-7wVxE2LNxaY6ZoxcWAmhiyrqaLZgf\",\n",
     "  \"object\": \"chat.completion\",\n",
     "  \"created\": 1694180212,\n",
     "  \"model\": \"gpt-35-turbo\",\n",
+    "  \"prompt_filter_results\": [\n",
+    "    {\n",
+    "      \"prompt_index\": 0,\n",
+    "      (content filter results removed for brevity)\n",
+    "    }\n",
+    "  ],\n",
     "  \"choices\": [\n",
     "    {\n",
     "      \"index\": 0,\n",
@@ -208,7 +136,7 @@
     "}\n",
     "```\n",
     "\n",
-    "Here's some more information about some of the data contained in that response.\n",
+    "Here's some information about some of the more interesting pieces of data contained in that response.\n",
     "\n",
     "Key | Description\n",
     "--- | ---\n",
@@ -227,7 +155,7 @@
    "source": [
     "## Summary\n",
     "\n",
-    "In this lab, we used the Azure OpenAI API directly to query information about our instance of the Azure OpenAI service and to send a prompt to an OpenAI model."
+    "In this lab, we used the Azure OpenAI API directly to send a prompt to an OpenAI model."
    ]
   },
   {
@@ -266,7 +194,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.11.5"
+   "version": "3.11.7"
   },
   "orig_nbformat": 4
  },