AIStream-Peelout/flow-forecast

Deep learning PyTorch library for time series forecasting, classification, and anomaly detection (originally for flood forecasting).

What it solves

Flow Forecast is a deep learning framework designed to simplify time series forecasting, classification, and anomaly detection. It provides a unified interface to implement and deploy state-of-the-art models, reducing the effort required to build end-to-end pipelines for temporal data.

How it works

The framework implements a wide variety of deep learning architectures specifically tuned for time series, including Transformers, Attention models, GRUs, LSTMs, and ODEs. It integrates with cloud providers like Google Cloud Platform and tracking tools like Weights and Biases to support the full lifecycle from training to model serving.

Who it’s for

It is built for data scientists and ML engineers who need to perform time series analysis, particularly those working with multivariate data or those specializing in environmental forecasting (such as river flow and flash flood prediction).

Highlights

  • Extensive Model Library: Supports a broad range of models including Informer, Transformer XL, DeepAR, TSMixer, and Crossformer.
  • End-to-End Capability: Covers the entire pipeline from training to model serving.
  • Interpretability: Includes metrics to help users understand model decisions.
  • Cloud Integration: Native support for Google Cloud Platform and Weights and Biases.

Related

  • Project
    LongxingTan/Time-series-prediction

    A TensorFlow-based time series package that provides a unified interface for prediction, classification, and anomaly detection using state-of-the-art deep learning models.

  • Project
    hmorimitsu/ptlflow

    A unified PyTorch Lightning framework for training and evaluating a wide collection of state-of-the-art optical flow estimation models.

  • Project
    mikekeith52/scalecast

    A time series forecasting library that provides a uniform interface for ML and deep learning models, simplifying model selection, validation, and pipeline creation.

  • Project
    sktime/pytorch-forecasting

    A PyTorch-based library for time series forecasting that provides a high-level API and state-of-the-art deep learning architectures like Temporal Fusion Transformers and N-BEATS.

  • Project
    facebookresearch/Kats

    A comprehensive toolkit for time series analysis that provides a unified framework for forecasting, anomaly detection, and feature extraction.