python - 回退策略不适用于 RASA 框架

Question

我想实施两阶段后备政策，我跟着这篇文章

这是 config.yml

language: en
pipeline:
  # other policies
  - name: DIETClassifier
    epochs: 100
    entity_recognition: False
  - name: ResponseSelector
    epochs: 100
  - name: FallbackClassifier
    threshold: 0.7
policies:
  - name: TEDPolicy
    max_history: 10
    epochs: 20
  - name: AugmentedMemoizationPolicy
    max_history: 6
  - name: RulePolicy

在 rules.yml 中添加规则 -

rule: Implementation of the Two-Stage-Fallback 
  steps: - intent: nlu_fallback 
  - action: action_two_stage_fallback 
  - active_loop: action_two_stage_fallback

和 domain.yml 中的响应

responses:
  utter_ask_rephrase:
  - text: I'm sorry, I didn't quite understand that. Could you rephrase? 

并且在意图信心 goodbye0.23 上，它正在执行再见而不是后备策略。

我错过了什么吗？

标签： pythonmachine-learningrasarasa-x

Answer 1

我需要提到 RulePolicy 的一些参数

- name: RulePolicy
    core_fallback_threshold: 0.3
    core_fallback_action_name: "action_default_fallback"
    enable_fallback_prediction: True

现在这正在工作