config.yml

language: zh

pipeline:
  - name: "rasa_chinese.nlu.tokenizers.lm_tokenizer.LanguageModelTokenizer"
    tokenizer_url: http://localhost:8001
    # Flag to check whether to split intents
    intent_tokenization_flag: false
    # Symbol on which intent should be split
    intent_split_symbol: "_"
  - name: "nlu.lm_featurizer.LanguageModelFeaturizer"
    model_name: "bert"
    # model_weights: "bert-base-chinese"
    # model_weights: "./bert-models/chinese_wwm_ext_pytorch"
    model_weights: hfl/chinese-bert-wwm-ext
    from_pt: true
    cache_dir: "./cache"
  - name: "nlu.recognizers_service_extractor.RecognizersServiceEntityExtractor"
    culture: 'zh-cn'
  - name: "RegexFeaturizer"
  - name: "LexicalSyntacticFeaturizer"
  - name: "CountVectorsFeaturizer"
  - name: "CountVectorsFeaturizer"
    analyzer: "char_wb"
    min_ngram: 1
    max_ngram: 7
  - name: "DIETClassifier"
    epochs: 100
  - name: "EntitySynonymMapper"

policies:
  - name: MemoizationPolicy
  - name: TEDPolicy
    max_history: 10
    epochs: 100
  - name: RulePolicy