TimEE: End-to-end Time Series Classification via In-Context Learning

TimEE is a 4.5M-parameter foundation model that classifies time series in a single forward pass using a few labeled examples — no training or fine-tuning required.

How to use:

1. Support CSV: Each row is one labeled time series. First column = class label, remaining columns = values.
2. Query CSV: Each row is one unlabeled time series. All columns = values.
3. Click Classify to see predictions and class probability distributions.

📄 Paper | 🐙 GitHub | 🤗 Model