| --- |
| title: TinyCast Forecaster |
| emoji: 📈 |
| colorFrom: gray |
| colorTo: blue |
| sdk: gradio |
| sdk_version: 6.25.0 |
| app_file: app.py |
| short_description: TinyCast zero-shot probabilistic time-series forecasting |
| python_version: "3.12" |
| startup_duration_timeout: 30m |
| --- |
| |
| # TinyCast Forecaster |
|
|
| A Gradio demo for [TinyCast](https://huggingface.co/raws-labs/tinycast), an attention-free, |
| 146,505-parameter time-series foundation model that forecasts unseen series zero-shot and |
| returns nine quantile forecasts. |
|
|
| ## How it works |
|
|
| 1. Upload a CSV file with a value column (or paste numeric values directly) |
| 2. Select the frequency/domain of your time series |
| 3. Choose a forecast horizon (1–512 steps) |
| 4. Get a probabilistic forecast plot with quantile confidence bands |
|
|
| The model replaces self-attention with dilated causal convolutions and a zero-parameter |
| normalized-periodogram phase prior, so periodicity is computed from the context instead |
| of learned. Every learned operation is a convolution, a matrix multiplication or a |
| normalization, so the model streams in constant memory and runs on CPU. |
|
|
| ## Model |
|
|
| - **Weights**: [raws-labs/tinycast](https://huggingface.co/raws-labs/tinycast) |
| - **Code**: [github.com/raws-labs/tinycast](https://github.com/raws-labs/tinycast) |
| - **Paper**: [arXiv:2608.15767](https://arxiv.org/abs/2608.15767) |
| - **License**: Apache-2.0 |
