Here's a detailed walkthrough of what you need to get started with the labs.
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.
Someone may have already set these up for you to use, however if you need to deploy your own you can follow these instructions.
- Check if you already have an Azure OpenAI service deployed. Go to the Azure Portal and in the top search bar, enter
openai. You'll then see Azure OpenAI listed under Services. Click on Azure OpenAI to see all deployed services.
On the next page, you should see a list of deployed services. If you have one and are able to access it, click on it. Otherwise, click the + Create button at the top of the page to create a new Azure OpenAI service.
- With a new or existing instance of Azure OpenAI deployed, click on the instance. On the main Overview page, you will see a Get Started section. Click on the Develop link.
On the Develop page you will see values for Key 1, Key 2, Location / Region and Endpoint. You will need the Key 1 and Endpoint details later on.
- Back at the Overview page in the Get Started section, click on Explore. This will launch the Azure AI Studio. Once in Azure AI Studio, click on the Deployments link on the left hand side.
- In the deployments section, we can make sure that we have one of each of the required models available.
NOTE: Please take extra care to make sure the correct version of each model is deployed.
You can see above that we have a completions model gpt-35-turbo with version 0613 and an embeddings model text-embedding-ada-002 with version 2. If you have both of these, then you're good to go. If not, click on the + Create new deployment link and follow the steps to create two deployments. Ensure that one model deployment uses text-embedding-ada-002 and the other uses a completions model such as gpt-35-turbo.
Make a note of both the deployment name and the model name for each of the two deployments.
We need to update a configuration file in this repo so that the labs are able to make use of your Azure OpenAI service. In the root of this repository, you will see a file named env.example. Make a copy of this file in the same location (the root of the repository) and rename the file to .env (don't forget the period at the beginning of the filename!)
Here's a detailed list of what you need to enter into the .env file and where to find the information you need to supply.
OPTIONAL: The OPENAI_API_TYPE value can be set to one of two values, depending on how you plan to authenticate to the Azure OpenAI service. To use an API key, set this to azure (default). To use Azure AD authentication, set this to azure_ad (currently not supported by this lab)
OPENAI_API_TYPE = "azure"
If you're using Azure AD authentication, the AZURE_OPENAI_API_KEY value can be left empty. It will be populated with an Azure AD token at runtime.
If you're using API keys, the AZURE_OPENAI_API_KEY is the Key 1 value you found on the Develop tab of the Overview page for the Azure OpenAI Service in the Azure portal.
AZURE_OPENAI_API_KEY = "5a8d1ea15ba00f1a833ab1ff245cdb0a"
The AZURE_OPENAI_ENDPOINT is the Endpoint value you found on the Develop tab of the Overview page for the Azure OpenAI Service in the Azure portal.
It will look similar to the below, ending with .openai.azure.com.
AZURE_OPENAI_ENDPOINT = "https://my-openaiservice.openai.azure.com/"
The OPENAI_API_VERSION is used for the Azure OpenAI API to determine which version of the API to use. The value below should be fine, but you can view available versions of the API at the following link: https://learn.microsoft.com/en-us/azure/cognitive-services/openai/reference#rest-api-versioning
OPENAI_API_VERSION = "2023-09-01-preview"
The next sections all relate to the models you have deployed in the Azure OpenAI service.
First is AZURE_OPENAI_COMPLETION_MODEL. This is the name of the completions model. It's likely to be gpt-35-turbo, but if you're using a different completions model, provide the name of the model here.
Note for this value, it's the name of the model, NOT the name of the deployment.
AZURE_OPENAI_COMPLETION_MODEL = "gpt-35-turbo"
The next two items are the deployments you have created using the Azure OpenAI service. First is the name of the deployment for the completions model and can be found in the Deployment name column of the Deployments page in Azure AI Studio.
AZURE_OPENAI_COMPLETION_DEPLOYMENT_NAME = "gpt35turbo"
The final value is the name of the deployment for the embeddings model and can be found in the Deployment name column of the Deployments page in Azure AI Studio.
AZURE_OPENAI_EMBEDDING_DEPLOYMENT_NAME = "embedding"
With all of the above updates to the .env file made, make sure you save the file and then you are ready to start the labs.
NOTE: The .gitignore file in this repo is configured to ignore the .env file, so the secrets such as the API key will not be uploaded to a public repo.
📣 Prompts