At its Azure Open Source Day 2023 today, Microsoft announced a handful of Azure AI open source capabilities to help developers build intelligent apps faster. These enhancements build on the Azure OpenAI Service capabilities that Microsoft revealed in January.

“AI is changing every industry and is top of mind for developers,” Microsoft Director David Dennis explains. “Most companies have leveraged AI to improve efficiency and costs. Large AI applications leveraging natural language processing (NLP), automatic speech recognition (ASR), and text-to-speech (TTS) are becoming prevalent, but what powers these applications is the underlying infrastructure optimized for large AI workloads. Azure is the best place to build AI workloads.”

Some of the new capabilities discussed today include:

Foundation Models in Azure Machine Learning. This new feature allows users to fine-tune and deploy pre-trained open source foundation models from Hugging Face throughout Azure ML.

Responsible AI Toolbox additions. Announced last week, the new Responsible AI Mitigations Library and the Fairlearn fairness assessment tool are open source tools designed to ease the adoption of responsible AI practices.

Vision Services. Vision Services are now available in public preview, helping developers improve content discoverability through automatic captioning, video summarization, background removal, image retrieval, and more. There are also new responsible AI controls for movement tracking, environment analysis, and real-time alerts.

You can learn more at the Azure Open Source Day website.