NVIDIA introduces Kumo Forecast for multivariate time-series predictions

First reported by Hugging Face at · Updated · 2 sources

The model combines pretrained time-series representations with historical data retrieval and tools for interpreting predictions. According to Unite.AI, it uses a MOMENT-1-large Transformer backbone with optional retrieval, is available under the Apache 2.0 license, and was previously known as NV-Tesseract Forecasting before NVIDIA announced the new name.

  • NVIDIA detailed the model in a Hugging Face blog post, according to Unite.AI.

Covered by 2 publishers within an hour of the first report: 1 news report and 1 from the company.

Reporting1

Unite.AI NVIDIA Announces Open Kumo Forecast Multivariate Time-Series Model · Jonas Reeve, Cognitive AI & AGI, AI Research Agent

From the company1

Hugging Face NVIDIA Kumo Forecast Combines Pretrained Time-Series Representations with Historical Pattern Retrieval and Interpretability

Related

Nvidia introduces RFD4-Proteina model for protein design
Nvidia, Anthropic Weigh In On AI Safeguards
T-Mobile completes AI-RAN call with Nokia and Nvidia
Nvidia publishes Cosmos3-H-Dreams model for surgical simulation

Topics Nvidia