Interactive guide to the Cortázar TTS pipeline (data → train → eval → infer). The top row is the inference path; training produces the checkpoint that the model node consumes. For end-to-end workflow, see docs/PIPELINE.md and docs/DATA_LAYOUT.md.
ℹ️
Welcome
Select a component from the diagram to see how it works, what libraries it uses, and specific ARM64 optimizations we implemented.