Yug: a foundation model for zero-shot probabilistic time-series forecasting
First Apache-2.0 Turkish-vertical TSFM fine-tune: Chronos-2 + LoRA r=32 + V4 router (MASE-aware adoption) + bucket isotonic calibration. GIFT-Eval submission 2026-05-27.
Fracast-0 is an 85K-parameter pretrained Time Series Foundation Model, small enough to load directly in a web page.
A systematic robustness testing tool for time series foundation models in edge cases.
Foundation model evaluation on pupillary light reflex time series. Biosignal analysis with modern TSFMs.
A grounded time-series forecasting framework that combines a language controller with specialist forecasting models. It turns natural-language requests into structured forecasting plans and evidence-grounded predictions, substantially outperforming general-purpose conversational language models used as direct forecasting baselines.
Paper's code of "How Effective Are Time Series Foundation Models for Epidemiological Data Analysis?"
MCP server for time-series foundation model (TSFM) forecasting: probabilistic forecasts from Chronos-2, TimesFM 2.5, Toto 2, TiRex-2 and IBM Granite models, or an accuracy-weighted ensemble.
Research and operations platform for statistically rigorous lottery forecasting experiments across six games.
InfraWatch: Time series foundation model anomaly detection for infrastructure metrics. Zero-config, maintenance-window aware.
To associate your repository with the tsfm topic, visit your repo's landing page and select "manage topics."