notice

This is unreleased documentation for Rasa Documentation Main/Unreleased version.
For the latest released documentation, see the latest version (3.x).

Version: Main/Unreleased

rasa.core.training.training

ActionFingerprint Objects

@dataclasses.dataclass
class ActionFingerprint()

Dataclass to represent an action fingerprint.

create_action_fingerprints

def create_action_fingerprints(
trackers: List["DialogueStateTracker"],
domain: "Domain") -> Dict[Text, ActionFingerprint]

Fingerprint each action using the events it created during train.

This allows us to emit warnings when the model is used if an action does things it hasn't done during training, or if rules are incomplete.

Arguments:

  • trackers - the list of trackers
  • domain - the domain

Returns:

a nested dictionary of action names and slots and active loops that this action sets