IBM releases SOTA Granite Time Series PatchTST-FM-r2 model with commercial-friendly license

IBM has released the SOTA Granite Time Series PatchTST-FM-r2 model, featuring a commercial-friendly open license and high-performance zero-shot forecasting. This model uses self-attention to capture long-range relationships and convolution for local temporal structures, allowing attention to focus on longer horizons. Its conformer blocks utilize alternating convolution kernel sizes of 3 and 5, in a repeating pattern {5, 5, 3, 3}. The pipeline generates future forecasts, including requested quantiles, without fine-tuning or task-specific model fitting.

Unlike previous models that often require fine-tuning, this IBM release offers zero-shot forecasting with a commercial-friendly license, a first for this level of performance.