Version: 3.x
rasa.engine.training.fingerprinting
Fingerprintable Objects
@runtime_checkable
class Fingerprintable(Protocol)
Interface that enforces training data can be fingerprinted.
fingerprint
def fingerprint() -> Text
Returns a unique stable fingerprint of the data.
calculate_fingerprint_key
def calculate_fingerprint_key(graph_component_class: Type[GraphComponent],
config: Dict[Text, Any],
inputs: Dict[Text, Fingerprintable]) -> Text
Calculates a fingerprint key that uniquely represents a single node's execution.
Arguments:
graph_component_class- The graph component class.config- The component config.inputs- The inputs as a mapping of parent node name to input value.
Returns:
The fingerprint key.