New platform and service allows enterprises and sovereign nations to build custom generative AI models with domain-specific knowledge.
NVIDIA has introduced the new NVIDIA AI Foundry, an end-to-end platform and service enabling enterprises and sovereign nations to build custom generative AI models, with domain-specific knowledge.
The platform includes state-of-the-art foundation models, including the latest Llama 3.1 models, which can be customized with NVIDIA NeMo, which provides tools for data curation, synthetic data generation, fine-tuning, model evaluation, and guardrails.
NVIDIA AI Foundry is powered by NVIDIA DGX Cloud, co-engineered with leading cloud providers and NVIDIA AI architecture.
The output of the Foundry is an NVIDIA NIM inference microservice, containing the custom model, can be easily deployed for production on the user’s platform of choice.
Get more information here.
Source: NVIDIA
Debbie Diamond Sarto is news editor at Animation World Network.
