---
type: "raw_signal"
id: "nvidia-kumo-forecast-combines-pretrained-time-series-representations-1ce161"
slug: "nvidia-kumo-forecast-combines-pretrained-time-series-representations-1ce161"
title: "NVIDIA Kumo Forecast Combines Pretrained Time-Series Representations with Historical Pattern Retrieval and Interpretability"
description: "NVIDIA has released Kumo Forecast, an Apache-2.0 multivariate time-series forecasting model and library."
retrieval_nugget: "NVIDIA has released Kumo Forecast, an Apache-2.0 multivariate time-series forecasting model and library."
observed_at: "2026-10-21"
record_date: "2026-10-21"
date_kind: "observed_at"
why_it_matters: "Source: https://huggingface.co/blog/nvidia/kumo-forecast"
novelty: "notable"
verification_level: "source-linked"
signal_type: "release"
source_platform: "huggingface.co"
topics: []
entities: ["NVIDIA","Kumo Forecast","Kumo-TS","MOMENT-1-large"]
related_patterns: []
novelty_label: "new_fact"
what_changed: "Adds detailed Kumo Forecast release information: its pretrained time-series model, historical-pattern retrieval, interpretability features, and Apache 2.0 availability."
prior_refs: ["memory:6e1add91-cae8-83e1-a751-b1c8c1d895a7"]
source_url: "https://huggingface.co/blog/nvidia/kumo-forecast"
source_urls: ["https://huggingface.co/blog/nvidia/kumo-forecast"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "signal:nvidia-kumo-forecast-combines-pretrained-time-series-representations-1ce161"
status: "published"
visuals: []
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/signals/nvidia-kumo-forecast-combines-pretrained-time-series-representations-1ce161/","markdown":"https://newruntime.com/signals/nvidia-kumo-forecast-combines-pretrained-time-series-representations-1ce161.md","json":"https://newruntime.com/signals/nvidia-kumo-forecast-combines-pretrained-time-series-representations-1ce161.json"}
---

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

## Retrieval answer

NVIDIA has released Kumo Forecast, an Apache-2.0 multivariate time-series forecasting model and library.

NVIDIA has released Kumo Forecast, an Apache-2.0 multivariate time-series forecasting model and library. It uses the approximately 340-million-parameter MOMENT-1-large encoder with a forecasting head, and offers an optional Domain-Aware Representation and Retrieval (DARR) mode.

DARR retrieves related historical patterns from a separate context dataset and blends them with a direct forecast. The library exposes a DataFrame-first interface, configurable horizons, autoregressive extension, and forecast-horizon explanations including lag and feature attributions. NVIDIA advises evaluating direct and retrieval-enhanced configurations on held-out windows for the intended operating conditions.

Source: https://huggingface.co/blog/nvidia/kumo-forecast

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- collection: [Raw Signals](https://newruntime.com/signals.json): Open every source-linked raw signal.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)