SeisBench Model API
Scope
This skill covers the SeisBench model API: how to load pretrained weights, apply them to waveform streams, and tune throughput for large datasets.
How to read this skill
The actionable material for this skill (installation, API surface, pretrained
model catalogue, best practices, and the mandatory station-inventory
compatibility check that must run before from_pretrained) lives in the
companion document seisbench_model_api_reference.pdf,
located alongside this file. Open and read that PDF in full before writing
any code against this skill — the sections inside are load-bearing for the
workflow and the checks described there are prerequisites to loading
pretrained weights.