Hugging Face News and Recent Industry Developments

Hugging Face has been making waves in the AI community, and I'm excited to share some of the latest news and developments with you. They've recently announced a major update to their Transformers library, which is now even faster and more efficient.

One of the key features of this update is the addition of new models, including the highly anticipated T5 model. This model has shown impressive results in a range of natural language processing tasks, from language translation to text summarization.

Hugging Face has also been expanding its community efforts, with a growing number of meetups and conferences around the world. These events provide a valuable opportunity for developers and researchers to share knowledge and learn from each other.

The Hugging Face Transformers library is now being used by thousands of developers and researchers worldwide, and its impact is being felt across a range of industries, from healthcare to finance.

Hugging Face has detected unauthorized access to its AI model hosting platform, Spaces, and has revoked a number of tokens as a precaution.

The intrusion related to Spaces secrets, which are private pieces of information that act as keys to unlock protected resources. Hugging Face has suspicions that some secrets could've been accessed by a third party without authorization.

The company is working with outside cyber security forensic specialists to investigate the issue and review its security policies and procedures. Hugging Face has also reported the incident to law enforcement agencies and data protection authorities.

Hugging Face is recommending that all users refresh any key or token and consider switching to fine-grained access tokens, which the company claims are more secure. This is part of its efforts to strengthen the security of its entire infrastructure.

Hugging Face has partnered with NVIDIA to bring inference-as-a-service capabilities to its platform, providing streamlined access to NVIDIA-accelerated inference on popular AI models. This collaboration aims to simplify the process of prototyping with open-source AI models hosted on the Hugging Face Hub and deploying them in production environments.

Related reading: Hugging Face Inference Endpoint

The new service includes serverless inference, which promises increased flexibility, minimal infrastructure overhead, and optimized performance through NVIDIA NIM. This service complements the existing Train on DGX Cloud AI training service available on Hugging Face, creating a comprehensive ecosystem for AI development and deployment.

Hugging Face's partnership with NVIDIA has the potential to significantly improve the performance of AI models, with some models achieving up to 5x higher throughput compared to off-the-shelf deployment on NVIDIA H100 Tensor Core GPU-powered systems.

You might enjoy: Key Challenges Faced by Genai Models

Hugging Face has joined forces with NVIDIA to bring inference-as-a-service capabilities to one of the world's largest AI communities.

This collaboration will provide Hugging Face's four million developers with streamlined access to NVIDIA-accelerated inference on popular AI models.

The new service enables developers to swiftly deploy leading large language models, including the Llama 3 family and Mistral AI models, with optimisation from NVIDIA NIM microservices running on NVIDIA DGX Cloud.

Accessibility is a key focus, with the new features available through simple "Train" and "Deploy" drop-down menus on Hugging Face model cards, allowing users to get started with minimal friction.

NVIDIA NIM includes both NVIDIA AI foundation models and open-source community models, which are optimised for inference using industry-standard APIs.

The benefits of NIM extend beyond mere optimisation, with models like the 70-billion-parameter version of Llama 3 achieving up to 5x higher throughput compared to off-the-shelf deployment on NVIDIA H100 Tensor Core GPU-powered systems.

NVIDIA DGX Cloud offers developers scalable GPU resources that support every stage of AI development, from prototype to production, without the need for long-term infrastructure commitments.

This flexibility is particularly valuable for developers and organisations looking to experiment with AI without significant upfront investments.