Hugging Face, an AI community specialized in machine learning, natural language processing, and deep learning, has recently announced the launch of a new product named Hugging Face Endpoints on Azure in partnership with tech giant Microsoft.
Clément Delangue, CEO and co-founder of Hugging Face, elaborated on the new product as a key to turning Hugging Face-developed Artificial Intelligence (AI) models into scalable solutions.
He further said that Hugging face’s mission is to make good machine learning accessible. The company is striving hard to assist every organization to develop high-quality ML-powered applications that will have a positive impact on businesses.
With Hugging Face Endpoints, the company has simplified the process of deploying state-of-the-art models and is excited to see what Azure users will build with them, said Delangue.
Hugging Face Endpoints will be available through Azure ML services that will allow users to activate Hugging Face services with some clicks of Microsoft Azure SDK code. After selecting the particular task type, users can integrate the Hugging Face model on any internal interface, such as application, website, or backend service.
Jeff Boudier, Hugging Face's product director, says that transformer models have changed the way companies build technology with ML. Endpoints will democratize transformers to Azure customers.
He further talked about adding support for tasks related to computer vision and speech including image classification and automatic speech recognition to current Hugging Face Endpoints in the next couple of weeks.
Hugging Face Endpoints has recently launched in Beta, and is already being utilized by large South African bank, Standard Bank. The model will be free during the whole preview period, wherein customers will only have to pay for the underlying Azure compute, however, customers will have to pay for the pricing model as per usage.