Google Introduces TimesFM-3 Forecasting Model
The model delivers zero-shot multivariate forecasting but drops its open license.
The model delivers zero-shot multivariate forecasting but drops its open license.
Google has introduced TimesFM-3, a forecasting model designed for time series data. The New Stack reports that the model delivers zero-shot multivariate forecasting and outperforms competing approaches on standard benchmarks. The release drops the open license that characterized earlier versions in the family. This change means the model cannot yet be used in professional settings despite its technical strengths. The New Stack article covers the model's development and its implications for software deployment and management in tech environments focused on at-scale operations.
639
1 post, first seen 4d ago
The model delivers zero-shot multivariate forecasting but drops its open license.
Google has introduced TimesFM-3, a forecasting model designed for time series data. The New Stack reports that the model delivers zero-shot multivariate forecasting and outperforms competing approaches on standard benchmarks. The release drops the open license that characterized earlier versions in the family. This change means the model cannot yet be used in professional settings despite its technical strengths. The New Stack article covers the model's development and its implications for software deployment and management in tech environments focused on at-scale operations.