“Developing and deploying AI applications at scale requires a robust and flexible platform that can handle the complex and diverse needs of modern businesses and enable them to build solutions based on their organizational data,” Microsoft said in announcing the updates.
Upcoming updates for Microsoft’s AI models and cloud platform
Microsoft said it offers serverless fine-tuning for Phi-3-mini and Phi-3-medium models to enable developers to quickly and easily customize models for cloud and edge scenarios without having to organize compute.
The company also announced updates to the Phi-3-mini AI model, including improvements to core quality, instruction tracking, and structured output. This will allow developers to create a more performant model at no additional cost.
Microsoft announced earlier this month that it will ship the latest models of its OpenAI (GPT-4o mini), Meta (Llama 3.1 405B), Mistral (Large 2) on Azure AI.
What are Phi-3 models?
In April, Microsoft introduced the Phi-3 family of small open models. The company claims that the Phi-3 models are its highest-performing and most cost-effective reduced language models (SLMs) available, and claims that they outperform models of the same and larger sizes.
“As developers look to tailor AI solutions to meet specific business needs and improve the quality of responses, fine-tuning a small model is a great alternative without sacrificing performance,” the company said, adding that developers can fine-tune Phi-3-mini and Phi-3-medium with their data to create more relevant AI experiences for their users, safely and cost-effectively.
