Chronos is a family of pretrained time series forecasting models based on language model architectures. A time series is transformed into a sequence of tokens via scaling and quantization, and a language model is trained on these tokens using the cross-entropy loss. Once trained, probabilistic forecasts are obtained by sampling multiple future trajectories given the historical context. Chronos models have been trained on a large corpus of publicly available time series data, as well as synthetic data generated using Gaussian processes.
Features
- The models in this repository are based on the T5 architecture
- Documentation available
- Zero-Shot Results
- The recommended way of using Chronos for production use cases is through AutoGluon
- Forecasting
- Examples available
- Pretraining, fine-tuning and evaluation
Project Samples
Categories
Machine LearningLicense
Apache License V2.0Follow Chronos Forecasting
You Might Also Like
Enterprise-Grade Monitoring - Zero Compromises
Managing a large, distributed IT environment demands more than basic monitoring. PRTG provides a unified view of your entire infrastructure - across sites, clouds, and hybrid setups. Advanced analytics, customizable dashboards, and granular access controls empower your team to detect issues early and respond fast. Automate reporting, ensure compliance, and scale effortlessly as your network grows. With PRTG, you get reliability, flexibility, and the insights you need to keep your business running at peak performance.
Rate This Project
Login To Rate This Project
User Reviews
Be the first to post a review of Chronos Forecasting!