C

Chronos T5 Mini

Developed by autogluon
Chronos is a family of pretrained time series forecasting models based on language model architectures, trained by converting time series into token sequences, supporting probabilistic forecasting.
Downloads 23.57k
Release Time : 5/14/2024

Model Overview

Chronos-T5 is a foundational model for time series forecasting, utilizing the T5 architecture. It converts time series into token sequences through quantization and scaling for training. The model supports sampling multiple future trajectories given historical context to obtain probabilistic forecasts.

Model Features

Probabilistic Forecasting
Obtains forecast distributions through multiple trajectory samplings, providing uncertainty estimates.
Pretrained Foundation Model
Pretrained on extensive public time series data and synthetic data, offering generalization capabilities.
Efficient Architecture
Utilizes an optimized T5 architecture with a vocabulary size of 4096 tokens for high parameter efficiency.
Multi-domain Applicability
Applicable to various time series forecasting scenarios without requiring domain-specific fine-tuning.

Model Capabilities

Time Series Forecasting
Probabilistic Forecasting
Multi-step Forecasting
Zero-shot Forecasting

Use Cases

Business Forecasting
Sales Forecasting
Predict future product sales volumes
Foot Traffic Forecasting
Predict future visitor numbers for malls or tourist attractions
Economics & Finance
Stock Price Prediction
Forecast trends in financial asset prices
Economic Indicator Forecasting
Predict macroeconomic indicators such as GDP and unemployment rates
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
© 2025AIbase