NVIDIA releases Kumo Relational, a pre-trained model for predictions on relational databases
AINVIDIA releases Kumo Relational, a pre-trained foundation model that predicts outcomes such as churn, demand, or fraud directly on relational databases. It uses up to 10,000 historical examples as in-context learning and needs no task-specific training or feature engineering.
AIWhy it matters.
The post explains how the model replaces hand-built feature pipelines on relational data, with benchmark results against supervised baselines and a clear note on its limits.
Source details
Hugging Face Blog · huggingface.coPublished · added here
Original title: NVIDIA Kumo Relational Outperforms Supervised Machine Learning on Relational Data, No Training Required