Skip to content
EVENT / TOPICHistorical

NVIDIA Unveils Kumo Forecast Time-Series Prediction Model

1 reports1 sources10/21/2026 updated

Current overview (not historical)

AI overview

NVIDIA has released Kumo Forecast, a time-series prediction model built on the MOMENT-1-large backbone. The model combines pretrained representations with historical pattern retrieval to enhance contextual forecasting. It features a configurable prediction horizon, an interpretability framework, and is released under the Apache 2.0 license. Developers can utilize the DataFrame-first API for direct prediction or fine-tuning on proprietary data.

Written with AI from the reporting · updated 10/21/2026

Latest developmentOct 21, 20:51
NVIDIA releases Kumo Forecast, a multi-variate time-series model combining pretrained representations with historical pattern retrieval
Explore the timeline, relations and evidence
Expand the detailed fact route and sources

Story route

A station-by-station route from the first public evidence onward; every station carries its own evidence and uncertainty.

  1. 1Unclassified fact stationOct 21, 00:00recorded Oct 21, 21:44single sourcerevised

    NVIDIA发布Kumo Forecast时间序列预测模型

    Read original source ↗ Hugging Face BlogReplay this moment in the explorer
    Evidence & revisions · 1 source records
    Subject：
    nvidia
    Action：
    发布
    Object：
    Kumo Forecast时间序列预测模型
    Occurred (fact side)：
    Oct 21, 00:00
    Hugging Face Blog ↗Report pageRevision 2 · currently revised
    Published (valid side)：
    Oct 21, 20:51
    First seen (known side)：
    Oct 21, 21:44
    Fact stationSingle source · one-sided claimRevised

    Copy may have been revised later; time replay is bounded. Layer labels group reading only, not causation between layers.

Uncertainty marks flag where to check the sources; they are not proven causation.

Report timeline

Explore every side through dated source reports.

Oct 21
  1. Hugging Face BlogFirst-party sourceRelated fact: single sourceSource report revised
    NVIDIA releases Kumo Forecast, a multi-variate time-series model combining pretrained representations with historical pattern retrieval

    Report first recorded Oct 21, 21:44

    NVIDIA Kumo Forecast is a multi-variate time-series forecasting model that combines a pretrained Transformer backbone with Domain-Aware Representation and Retrieval (DARR). Built on the MOMENT-1-large encoder with approximately 340 million parameters, it offers configurable forecast horizons and an optional retrieval pathway that blends direct model predictions with patterns retrieved from historical data. The library provides a DataFrame-first API and an interpretability framework producing horizon-specific explanations.

    Read original source ↗

Event discussion

Reader opinions are hosted on GitHub Discussions; posting requires a GitHub account. Comments are not verified facts.

Open GitHub Discussions