Chat-TS training datasets

This dataset is intended to train reasoning.

The dataset consists of real-world time-series with synthetic text.

If you use this work please cite:

```bash
@misc{quinlan2025chattsenhancingmultimodalreasoning,
      title={Chat-TS: Enhancing Multi-Modal Reasoning Over Time-Series and Natural Language Data}, 
      author={Paul Quinlan and Qingguo Li and Xiaodan Zhu},
      year={2025},
      eprint={2503.10883},
      archivePrefix={arXiv},
      primaryClass={cs.AI},
      url={https://arxiv.org/abs/2503.10883}, 
}