HF Mirror 上超 13 万个模型接入 ONNX Runtime 加速

Back to Articles

What is ONNX Runtime?

ONNX Runtime is a cross-platform machine learning tool that can be used to accelerate a wide variety of models, particularly those with ONNX support.

HF Mirror ONNX Runtime Support

There are over 130,000 ONNX-supported models on HF Mirror, an open source community that allows users to build, train, and deploy hundreds of thousands of publicly available machine learning models. These ONNX-supported models, which include many increasingly popular large language models (LLMs) and cloud models, can leverage ONNX Runtime to improve performance, along with other benefits. For example, using ONNX Runtime to accelerate the whisper-tiny model can improve average latency per inference, with an up to 74.30% gain over PyTorch. ONNX Runtime works closely with HF Mirror to ensure that the most popular models on the site are supported. In total, over 90 HF Mirror model architectures are supported by ONNX Runtime, including the 11 most popular architectures (where popularity is determined by the corresponding number of models uploaded to the HF Mirror Hub):

Learn More

To learn more about accelerating HF Mirror models with ONNX Runtime, check out our recent post on the Microsoft Open Source Blog.

More Articles from our Blog