Open Source Chronos-t5-large Time Series Forecasting Model - Free Implementation of Probabilistic Forecasting Applications

Chronos T5 Large

Developed by autogluon

Chronos is a family of pretrained time series forecasting models based on language model architecture, which supports probabilistic forecasting by converting time series into token sequences for training.

Climate Model

Transformers

Open Source License:Apache-2.0 #Time Series Forecasting #Probabilistic Forecasting #Language Model Architecture

Downloads 59.18k

Release Time : 5/14/2024

Model Overview

Chronos-T5 is a time series forecasting model based on the T5 architecture. It converts time series into token sequences through scaling and quantization, trained using cross-entropy loss, and capable of generating probabilistic forecasts with multiple future trajectories.

Model Features

Probabilistic Forecasting

Generates probabilistic forecasts by sampling multiple future trajectories, providing prediction distributions rather than single-point estimates.

Pretrained Model

Pretrained on extensive public time series data and synthetic data, ensuring broad applicability.

Efficient Inference

Compared to original models of the same size, the Chronos-Bolt series achieves up to 250x speed improvement and 20x memory efficiency.

Model Capabilities

Time Series Forecasting

Probabilistic Forecasting

Multi-Step Forecasting

Use Cases

Business Forecasting

Airline Passenger Forecasting

Predict future airline passenger numbers

Provides median forecasts and 80% prediction intervals for the next 12 months

Financial Analysis

Stock Price Prediction

Predict future stock price trends

🚀 Chronos-T5 (Large)

Chronos is a family of pretrained time series forecasting models based on language model architectures. It transforms time series into token sequences, trains a language model on these tokens, and then obtains probabilistic forecasts by sampling future trajectories. The models are trained on extensive public time - series data and synthetic data.

🚀 Quick Start

Update Feb 14, 2025: Chronos - Bolt & original Chronos models are now available on Amazon SageMaker JumpStart! Check out the tutorial notebook to learn how to deploy Chronos endpoints for production use in a few lines of code.
Update Nov 27, 2024: We have released Chronos - Bolt⚡️ models that are more accurate (5% lower error), up to 250 times faster and 20 times more memory - efficient than the original Chronos models of the same size. Check out the new models here.

For details on Chronos models, training data and procedures, and experimental results, please refer to the paper Chronos: Learning the Language of Time Series.

Fig. 1: High - level depiction of Chronos. (Left) The input time series is scaled and quantized to obtain a sequence of tokens. (Center) The tokens are fed into a language model which may either be an encoder - decoder or a decoder - only model. The model is trained using the cross - entropy loss. (Right) During inference, we autoregressively sample tokens from the model and map them back to numerical values. Multiple trajectories are sampled to obtain a predictive distribution.

✨ Features

Chronos models are based on language model architectures. They transform time series into token sequences, train with cross - entropy loss, and obtain probabilistic forecasts by sampling. They are trained on large - scale public time - series data and synthetic data.

📦 Installation

To perform inference with Chronos models, install the package in the GitHub companion repo by running:

pip install git+https://github.com/amazon-science/chronos-forecasting.git

💻 Usage Examples

Basic Usage

import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
import torch
from chronos import ChronosPipeline

pipeline = ChronosPipeline.from_pretrained(
  "amazon/chronos-t5-large",
  device_map="cuda",
  torch_dtype=torch.bfloat16,
)

df = pd.read_csv("https://raw.githubusercontent.com/AileenNielsen/TimeSeriesAnalysisWithPython/master/data/AirPassengers.csv")

# context must be either a 1D tensor, a list of 1D tensors,
# or a left-padded 2D tensor with batch as the first dimension
context = torch.tensor(df["#Passengers"])
prediction_length = 12
forecast = pipeline.predict(context, prediction_length)  # shape [num_series, num_samples, prediction_length]

# visualize the forecast
forecast_index = range(len(df), len(df) + prediction_length)
low, median, high = np.quantile(forecast[0].numpy(), [0.1, 0.5, 0.9], axis=0)

plt.figure(figsize=(8, 4))
plt.plot(df["#Passengers"], color="royalblue", label="historical data")
plt.plot(forecast_index, median, color="tomato", label="median forecast")
plt.fill_between(forecast_index, low, high, color="tomato", alpha=0.3, label="80% prediction interval")
plt.legend()
plt.grid()
plt.show()

📚 Documentation

The models in this repository are based on the T5 architecture. The only difference is in the vocabulary size: Chronos - T5 models use 4096 different tokens, compared to 32128 of the original T5 models, resulting in fewer parameters.

Property	Details
Model Type	The models are based on the T5 architecture with a different vocabulary size.
Training Data	Trained on a large corpus of publicly available time series data and synthetic data generated using Gaussian processes.

Model	Parameters	Based on
chronos-t5-tiny	8M	t5-efficient-tiny
chronos-t5-mini	20M	t5-efficient-mini
chronos-t5-small	46M	t5-efficient-small
chronos-t5-base	200M	t5-efficient-base
chronos-t5-large	710M	t5-efficient-large

📄 License

This project is licensed under the Apache - 2.0 License.

📚 Citation

If you find Chronos models useful for your research, please consider citing the associated paper:

@article{ansari2024chronos,
  author  = {Ansari, Abdul Fatir and Stella, Lorenzo and Turkmen, Caner and Zhang, Xiyuan, and Mercado, Pedro and Shen, Huibin and Shchur, Oleksandr and Rangapuram, Syama Syndar and Pineda Arango, Sebastian and Kapoor, Shubham and Zschiegner, Jasper and Maddix, Danielle C. and Mahoney, Michael W. and Torkkola, Kari and Gordon Wilson, Andrew and Bohlke - Schneider, Michael and Wang, Yuyang},
  title   = {Chronos: Learning the Language of Time Series},
  journal = {arXiv preprint arXiv:2403.07815},
  year    = {2024}
}

🔧 Security

See CONTRIBUTING for more information.

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご