SklearnIntentClassifier rasa 3.0

Rasa Open Source
1

Help me fix this error. Thanks

SklearnIntentClassifier rasa 3.0

raise GraphSchemaValidationException(rasa.engine.exceptions.GraphSchemaValidationException: 1 components are missing required components which have to run before themselves: The following components require a DenseFeaturizer: train_SklearnIntentClassifier5. Please add the required components to your model configuration.

file config language: vi pipeline:

  • name: WhitespaceTokenizer
  • name: RegexFeaturizer
  • name: LexicalSyntacticFeaturizer
  • name: CountVectorsFeaturizer
  • name: CountVectorsFeaturizer analyzer: char_wb min_ngram: 1 max_ngram: 4
  • name: SklearnIntentClassifier C: [1, 2, 5, 10, 20, 100] kernel: “linear” epochs: 100
  • name: EntitySynonymMapper
  • name: ResponseSelector epochs: 100
1 Like
2

The error message states that you need to include a dense featurizer in your pipeline. It’s also documented here.

You can find a description of the featurizers here followed by a list of the available featurizers. The two featurizers in your pipeline are both sparse featurizers.

I would try adding the LanguageModelFeaturizer to your pipeline.

- name: LanguageModelFeaturizer
  model_name: "bert"
  model_weights: "sentence-transformers/all-MiniLM-L6-v2"
1 Like
3

Thanks.

4

pipeline:

  • name: “SpacyNLP”

    model: “vi_core_news_lg” case_sensitive: False

  • name: “SpacyTokenizer”

  • name: “SpacyFeaturizer” pooling: mean

  • name: RegexFeaturizer

  • name: LexicalSyntacticFeaturizer

  • name: CountVectorsFeaturizer

  • name: CountVectorsFeaturizer

    analyzer: “char_wb”

    min_ngram: 1

    max_ngram: 4

  • name: “CRFEntityExtractor”

  • name: “classifiers.sklearn_intent_classifier.SklearnIntentClassifier”

    C: [1, 2, 5, 10, 20, 100]

    kernel: “linear”

and it works.

5

Hello, I have some problem with pipline when I try to use Underthesea tokenizer for my Vietnamese chatbot.

language: “vi”

pipeline:

  • name: “components.tokenizer.UndertheseaTokenizer”
  • name: “RegexFeaturizer”
  • name: “LexicalSyntacticFeaturizer”
  • name: “CountVectorsFeaturizer” analyzer: “char_wb” min_ngram: 1 max_ngram: 4
  • name: “CountVectorsFeaturizer”
  • name: “DIETClassifier” epochs: 100
  • name: “EntitySynonymMapper”
  • name: “ResponseSelector” epochs: 100
  • name: “FallbackClassifier” threshold: 0.3 ambiguity_threshold: 0.1 Error: rasa.engine.exceptions.GraphSchemaValidationException: 8 components are missing required components which have to run before themselves: The following components require a Tokenizer: run_CountVectorsFeaturizer3, run_CountVectorsFeaturizer4, run_LexicalSyntacticFeaturizer2, run_RegexFeaturizer1, train_CountVectorsFeaturizer3, train_CountVectorsFeaturizer4, train_LexicalSyntacticFeaturizer2, train_RegexFeaturizer1. Please add the required components to your model configuration.

Could you please help me? Can you contact me via email for easier discussion? ngohaiyen_t67@hus.edu.vn