Policies

Configuring Policies

The rasa.core.policies.Policy class decides which action to take at every step in the conversation.

There are different policies to choose from, and you can include multiple policies in a single rasa.core.agent.Agent.

Max History

One important hyperparameter for Rasa Core policies is the max_history. This controls how much dialogue history the model looks at to decide which action to take next.

policies:
  - name: "KerasPolicy"
    featurizer:
    - name: MaxHistoryTrackerFeaturizer
      max_history: 5
      state_featurizer:
        - name: BinarySingleStateFeaturizer
  - name: "MemoizationPolicy"
    max_history: 5
  - name: "FallbackPolicy"
    nlu_threshold: 0.4
    core_threshold: 0.3
    fallback_action_name: "my_fallback_action"
  - name: "path.to.your.policy.class"
    arg1: "..."

Data Augmentation

When you train a model, by default Rasa Core will create longer stories by randomly gluing together the ones in your stories files.

You can alter this behaviour with the --augmentation flag. The augmentation_factor determines how many augmented stories are subsampled during training.

Example:

--augmentation 0

Action Selection

At every turn, each policy defined in your configuration will predict a next action with a certain confidence level. The bot’s next action is then decided by the policy that predicts with the highest confidence.

Keras Policy

The KerasPolicy uses a neural network implemented in Keras to select the next action.

def model_architecture(self, input_shape: Tuple[int, int], output_shape: Tuple[int, Optional[int]]) -> tf.keras.models.Sequential:
    from tensorflow.keras.models import Sequential
    from tensorflow.keras.layers import (
        Masking,
        LSTM,
        Dense,
        TimeDistributed,
        Activation,
    )
    model = Sequential()
    ...
    model.compile(
        loss="categorical_crossentropy", optimizer="rmsprop", metrics=["accuracy"]
    )
    return model

Mapping Policy

The MappingPolicy can be used to directly map intents to actions.

intents:
 - ask_is_bot:
     triggers: action_is_bot

Memoization Policy

The MemoizationPolicy just memorizes the conversations in your training data. It predicts the next action with confidence 1.0 if this exact conversation exists in the training data.

Fallback Policy

The FallbackPolicy invokes a fallback action if certain confidence thresholds are not met.

Configuration:

policies:
  - name: "FallbackPolicy"
    nlu_threshold: 0.3
    ambiguity_threshold: 0.1
    core_threshold: 0.3
    fallback_action_name: 'action_default_fallback'

Two-Stage Fallback Policy

The TwoStageFallbackPolicy handles low NLU confidence in multiple stages.

Configuration:

policies:
  - name: TwoStageFallbackPolicy
    nlu_threshold: 0.3
    ambiguity_threshold: 0.1
    core_threshold: 0.3
    fallback_core_action_name: "action_default_fallback"
    fallback_nlu_action_name: "action_default_fallback"
    deny_suggestion_intent_name: "out_of_scope"

Form Policy

The FormPolicy is an extension of the MemoizationPolicy which handles the filling of forms.