Attempting to change the descriptions to use policy

config.yml

@@ -50,3 +50,4 @@ policies:
      epochs: 120
      learning_rate: 0.02
      constrain_similarities: true
+   - name: pipeline.product_description_policy.ProductDescriptionPolicy

pipeline/product_description_policy.py

@@ -0,0 +1,69 @@
+from typing import Dict, Text, Any, Optional
+
+from rasa.core.policies.policy import Policy
+from rasa.core.policies.policy import PolicyPrediction
+from rasa.shared.core.domain import Domain
+from rasa.shared.core.trackers import DialogueStateTracker
+from rasa.core.featurizers.precomputation import MessageContainerForCoreFeaturization
+from rasa.engine.storage.resource import Resource
+from rasa.engine.graph import ExecutionContext, GraphComponent
+from rasa.engine.storage.storage import ModelStorage
+from rasa.engine.recipes.default_recipe import DefaultV1Recipe
+from rasa.core.constants import (
+    POLICY_PRIORITY,
+    POLICY_MAX_HISTORY,
+)
+
+@DefaultV1Recipe.register(
+    [DefaultV1Recipe.ComponentType.POLICY_WITHOUT_END_TO_END_SUPPORT], is_trainable=False
+)
+class ProductDescriptionPolicy(Policy):
+    def __init__(
+        self,
+        config: Dict[Text, Any],
+        name: Text,
+        model_storage: ModelStorage,
+        resource: Resource,
+    ) -> None:
+        super().__init__(config, name, model_storage, resource)
+        self.config = config
+        self.name = name
+        self.model_storage = model_storage
+        self.resource = resource
+
+    @classmethod
+    def create(
+        cls,
+        config: Dict[Text, Any],
+        model_storage: ModelStorage,
+        resource: Resource,
+        execution_context: ExecutionContext,
+    ) -> GraphComponent:
+        return cls(config, execution_context.node_name, model_storage, resource)
+
+    @staticmethod
+    def get_default_config() -> Dict[Text, Any]:
+        return {
+            # Determines the importance of policies, higher values take precedence
+            POLICY_PRIORITY: 0.0,
+            POLICY_MAX_HISTORY: 10,
+        }
+
+    def predict_action_probabilities(
+        self,
+        tracker: DialogueStateTracker,
+        domain: Domain,
+        rule_only_data: Optional[Dict[Text, Any]] = None,
+        precomputations: Optional[MessageContainerForCoreFeaturization] = None,
+        **kwargs: Any,
+    ) -> PolicyPrediction:
+        # check if we are in an active loop
+        if tracker.active_loop:
+            print("ProductDescriptionPolicy: Active loop detected")
+        # print the latest intent detected
+        print("ProductDescriptionPolicy: Latest intent detected: ", tracker.latest_message.get("intent", {}).get("name"))
+        entites = tracker.latest_message.get("entities", [])
+        if any(entity["entity"] == "product" for entity in entites):
+            print("ProductDescriptionPolicy: Product entity detected")
+            print("ProductDescriptionPolicy: Product entity value: ", next(entity["value"] for entity in entites if entity["entity"] == "product"))
+        return PolicyPrediction([0.0 for _ in range(domain.num_actions)], "ProductDescriptionPolicy")