25 repos
Machine learning models and research implementations for predicting time series data using transformer architectures and related deep learning techniques. The cluster centers on foundational transformer-based forecasting models like Lag-Llama and PatchTST, alongside IBM's Granite time series models, with supporting research codebases and experimental variants. Developers and researchers exploring this area will find model implementations, fine-tuning frameworks, and benchmarking tools for applying modern neural architectures to temporal prediction problems.