Similar to Open AI, Azure Open AI Service is a cloud-based platform provided by Microsoft Azure that allows developers to integrate artificial intelligence capabilities into their applications. It offers a range of AI services such as natural language processing, speech recognition, image recognition, and more. With Azure Open AI Service, developers can leverage advanced AI models and APIs to enhance their applications with intelligent features.
Benefits:
- Provides pre-trained models
- Model Customisation Capabilities
- Tools to implement responsible AI
- Enterprise grade security
Who fits where?
How to use Azure OpenAI
- Register at aka.ms/oaiapply (It takes appx 24 hours for your account to get activated)
- Once you receive an email from Azure that your registration is completed then you can create instance in Azure (You must have an active azure account)
3. Use Azure OpenAI Studio https://oai.azure.com, On this page, you can explore task like summarising, GenAI, and classification texts, and there are resources for learning
On the left-hand corner, you will find Playground, where you can find two services:
- Completion Service: designed to generate text based on a given prompt or context.
- Chat: specifically tailored for engaging in open-ended conversations with the OpenAI models. It allows users to have a natural dialogue with the model, asking questions, answering prompts, and following the flow of the conversation.
There is another section called Management, where:
- Deployment: allows you to create, manage, and monitor your OpenAI deployments.
- Models: allows you to view and manage your OpenAI models.
- File Management: allows you to upload, download, and manage files that are associated with your OpenAI models.
Let’s explore the deployment section:
Here is a list of all the models available for deployment instantly.
Once you deploy the model, it will appear in the deployment section.
Now, when you visit the completion section, you will see the deployed model, and you can test it. You can see the drop-down section has all the deployed models, and on the right-hand section there will be all the parameters available to tweak the model. You can tune the model according to your needs.
Click here to learn what each parameter does
The second drop-down is for what you want to use the mode for.
In this example, we have selected Generate an Email. You will get a pre-written prompt, which you can change according to your usage. At the bottom, you can see the number of tokens used for this activity.
In another example, we can see that we selected Natural Language to SQL where we are getting a suggestion that this task works best with this model
Some FAQs which can be useful for us
How can I customize or automate the index creation process?
You can prepare the index yourself using a script provided on GitHub. Using this script will create an Azure AI Search index with all the information needed to better leverage your data, with your documents broken down into manageable chunks. Please see the README file with the data preparation code for details on how to run it.
How can I update my index?
You can schedule an automatic index refresh, or upload additional data to your Azure Blob Container and use it as your data source when you create a new index. The new index will include all of the data in your container.
Is there a token limit on the system message?
Yes, the token limit on the system message is 400. If the system message is more than 400 tokens, the rest of the tokens beyond the first 400 will be ignored. This limitation only applies to the Azure OpenAI on your data feature.
How can I add vector embeddings when indexing my data?
When you select “Azure Blob Container”, “Azure AI Search”, or “Upload files” as the data source, you can also select an Ada embedding model deployment to use when ingesting your data. This will create an Azure AI Search index with vector embeddings.
Thank you for reading! Let’s connect on LinkedIn.
