Rasa
環境構築
pip install rasa
pip install ginza ja-ginza
mkdir Rasa
cd Rasa
rasa init --no-prompt
日本語設定
config.yml
recipe: default.v1
assistant_id: placeholder_default
language: ja
pipeline:
- name: SpacyNLP
model: 'ja_ginza'
- name: SpacyTokenizer
- name: SpacyFeaturizer
- name: SpacyEntityExtractor
- name: RegexFeaturizer
- name: LexicalSyntacticFeaturizer
- name: CountVectorsFeaturizer
- name: CountVectorsFeaturizer
analyzer: "char_wb"
min_ngram: 1
max_ngram: 4
- name: DIETClassifier
epochs: 100
- name: EntitySynonymMapper
- name: ResponseSelector
epochs: 100
- name: FallbackClassifier
threshold: 0.7
policies:
- name: MemoizationPolicy
- name: RulePolicy
- name: TEDPolicy
max_history: 5
epochs: 100
設定追加
- domain.yml
- 「intents」に追加
- nlu.yml
- 「intent」を追加、学習させる例文を「examples」に入れる (日本語でも英語でも可)
- stories.yml
- 「rule」を追加、返答する内容等を「steps」に入れる
運用
訓練
rasa train
実行
rasa shell
参考文献
https://qiita.com/Zect/items/e341e43fa9cf98529942
https://tech-blog.optim.co.jp/entry/2021/11/17/100000