Blog Post

AI - Azure AI services Blog
3 MIN READ

Fine Tune GPT-4o on Azure OpenAI Service

AliciaFrame's avatar
AliciaFrame
Icon for Microsoft rankMicrosoft
Aug 28, 2024

Get excited - you can now fine tune GPT-4o using the Azure OpenAI Service!
 

We're thrilled to announce the public preview of fine-tuning for GPT-4o on Azure. After a successful private preview, GPT-4o is now available to all of our Azure OpenAI customers, offering unparalleled customization and performance in Azure OpenAI Service. 

 

Why fine-tuning matters 

 

Fine-tuning is a powerful tool that allows you to tailor our advanced models to your specific needs. Whether you’re looking to enhance the accuracy of responses, ensure outputs align with your brand voice, reduce token consumption or latency, or optimize the model for a particular use case, fine-tuning allows you to customize best-in-class models with your own proprietary data.  

 

GPT-4o: Make a great model even better with your own training data

 

GPT-4o offers the same performance as GPT-4 Turbo, but improved efficiency – and the best performance on non-English language content of any OpenAI model. With the launch of fine tuning for GPT-4o, you now have the ability to customize it for your unique needs. Fine-tuning GPT-4o enables developers to train the model on domain-specific data, creating outputs that are more relevant, accurate, and contextually appropriate. 

 

 

This release marks a significant milestone for Azure OpenAI Service, as it allows you to build highly specialized models that drive better outcomes, use fewer tokens with greater accuracy, and create truly differentiated models to support your use cases. 

    

Fine tuning capabilities 

 

Today, we're announcing the availability of text-to-text fine tuning for GPT-4o. In addition to basic customization, we support advanced features to help you create customized models for your needs: 

   

  • Tool Calling: Include function and tool calls in your training data to empower your custom models to do even more! 
  • Continuous Fine Tuning: Fine tune previously fine-tuned models with new, or additional, data to update or improve accuracy 
  • Deployable Snapshots: No need to worry about overfitting – you can now deploy snapshots, preserved at each epoch, and evaluate them alongside your final model 
  • Built in Safety: GPT-4, 4o, and 4o mini models have automatic guardrails in place to ensure that your fine-tuned models are not capable of generating harmful content.

GPT-4o is available to customers using Azure OpenAI resources in North Central US and Sweden Central. Stay tuned as we add support in additional regions.

   

Lowering prices to make experimentation accessible

  

We’ve heard your feedback about the costs of fine-tuning and hosting models. To make it easier for you to experiment and deploy fine-tuned models, we’ve updated our pricing structure to: 

These changes make experimentation easier (and less expensive!) than ever before. You can find the updated pricing for Fine tuning models on the Azure OpenAI Service - Pricing | Microsoft Azure  

   

Get started today! 

Whether you’re new to fine-tuning or an experienced developer, getting started with Azure OpenAI Service has never been easier. Fine-tuning is available through both Azure OpenAI Studio and Azure AI Studio, providing a user-friendly interface for those who prefer a GUI and robust APIs for advanced users. 

 

Ready to get started?  

 

 

 

Updated Aug 28, 2024
Version 2.0
  • ilik0's avatar
    ilik0
    Brass Contributor

    I have a question. If I build a specialized model and run it from Azure OpenAI how do I make sure the users using it pays at least what it costs me to run it?

    Is Azure APIM the only possible way?

  • Bitime's avatar
    Bitime
    Copper Contributor

    great news! is it possible to use text+image-to-text to fine tune GPT-4o?