| language: python | |
| license: custom | |
| library_name: transformers | |
| tags: | |
| - time-series | |
| - forecasting | |
| - foundation-model | |
| - transformer | |
| - decoder-only | |
| # EpochsFM-TF Model Card | |
| **Release Date:** September 2025 | |
| ## Model Details | |
| **Model Type:** Time series forecasting | |
| **Architecture:** Decoder-only transformer | |
| **Parameters:** 200M | |
| **Input:** Patch-based time series | |
| ## Intended Use | |
| - Time series forecasting tasks | |
| - Demand prediction | |
| - Financial modeling | |
| - Resource planning | |
| - Research and educational purposes | |
| ## Limitations | |
| - Designed for univariate time series (single target) | |
| - Requires sufficient context length for accurate forecasts | |
| - Performance varies across different domains | |
| ## Licensing | |
| Non-commercial use is permitted. Commercial deployments require explicit licensing. Contact inquiry@sagea.space. |