Learn how to run TimeGPT in a distributed manner on Spark for scalable forecasting and cross-validation.
nixtla
library (and any dependencies) are installed on all worker nodes to ensure consistent execution across the cluster.1. Installation
fugue
with Spark support:nixtla
library installed as well.2. Load Data
3. Initialize Spark
4. Use TimeGPT on Spark
Create a NixtlaClient Instance
base_url
parameter:Forecast
model="azureai"
.timegpt-1
(default) and timegpt-1-long-horizon
.
For long horizon forecasting, see
this tutorial.Cross-Validation
5. Stop Spark