Hugging Face pitches HUGS as an alternative to Nvidia's NIM for open models

Hugging Face launched HUGS, a viable alternative to Nvidia’s NIM, offering streamlined deployment of generative AI models with significant cost savings for enterprises.

Unlike Nvidia's NIM, which requires enterprise clients to pay $1 per hour per container, Hugging Face provides HUGS at a competitive $1 per hour, potentially enhancing savings.

HUGS utilizes open-source libraries like Text Generation Inference and Transformers, enabling efficient model inference on Nvidia and AMD GPUs, with plans for further hardware support.

Developers have shown excitement for HUGS due to its zero-configuration setup, allowing quick access to foundational models via preconfigured containers, a significant improvement over traditional frameworks.

Read at InfoWorld