We present TimeFound, an encoder-decoder transformer-based time series foundation model for out-of-the-box zero-shot forecasting. To handle time series data from various domains, TimeFound employs a multi-resolution patching strategy to capture complex temporal patterns at multiple scales. We pre-train our model with two sizes (200M and 710M parameters) on a large time-series corpus comprising both real-world and synthetic datasets. Over a collection of unseen datasets across diverse domains and forecasting horizons, our empirical evaluations suggest that TimeFound can achieve superior or competitive zero-shot forecasting performance, compared to state-of-the-art time series foundation models.
@article{arxiv.2503.04118,
title = {TimeFound: A Foundation Model for Time Series Forecasting},
author = {Congxi Xiao and Jingbo Zhou and Yixiong Xiao and Xinjiang Lu and Le Zhang and Hui Xiong},
journal= {arXiv preprint arXiv:2503.04118},
year = {2025}
}