espnet3.demo.runtime.run_inference

source

DemoRuntime, , ui_names: List[str], ui_values: List[Any], output_names: List[str]) → List[Any]

Run a single inference pass and map outputs for the UI.

This helper builds a one-item dataset from the UI inputs and then calls the configured runner (or the model directly if no runner is configured).

• Parameters:
  • runtime (DemoRuntime) – Demo runtime built by build_runtime().
  • ui_names (List *[*str ]) – UI input names (component names).
  • ui_values (List *[*Any ]) – UI input values (aligned with ui_names).
  • output_names (List *[*str ]) – UI output names to return values for.
• Returns: Output values aligned with output_names.
• Return type: List[Any]
• Raises:
  • ValueError – If output mapping is missing/mismatched for multi-output UIs.
  • RuntimeError – If no runner is configured and the model is not callable.

Example

>>> run_inference(
...     runtime,
...     ui_names=["speech"],
...     ui_values=["sample.wav"],
...     output_names=["text"],
... )