Usage guide
Setup, environment variables, and Python/JavaScript/Go usage examples.
TurnDetector
End-of-turn detector: decides when the caller has finished speaking. One class selected bymodel: "namo" (the default), "namo-inference",
"echo-small", or "echo-large". Which extra params apply depends on the
model (see __init__).
Constructor
Which detector to use:
"namo" (the default), "namo-inference", "echo-small", or "echo-large".End-of-turn probability cutoff, 0.0–1.0. Applies to every model. Default 0.7.
BCP-47 language hint. Only used by the
"namo" model.Inference service address. Inference models only.
Inference service auth token. Inference models only.
Inference service base URL. Inference models only.
aclose
cleanup
detect_end_of_utterance
The conversation context used to evaluate the utterance.
Optional probability threshold overriding the configured one.
True if the end-of-utterance probability meets or exceeds the threshold, otherwise False.emit
The event to emit.
get_eou_probability
0.5 when no result is available or the value cannot be parsed.
The conversation context used to evaluate the utterance.
The end-of-utterance probability in the range 0.0 to 1.0.
off
The event the handler was registered for.
The handler to remove.
on
callback is omitted.
The event to listen for.
The handler to invoke when the event is emitted. If
None, a decorator is returned that registers the decorated function.The registered callback when
callback is provided, otherwise a decorator that registers and returns the function it wraps.set_threshold
The new probability threshold.