target_extraction.allen.predictors package¶
Submodules¶
target_extraction.allen.predictors.target_sentiment module¶
target_extraction.allen.predictors.target_tagger module¶
-
class
target_extraction.allen.predictors.target_tagger.
TargetTaggerPredictor
(model, dataset_reader, language='en_core_web_sm', pos_tags=True, fine_grained_tags=False)[source]¶ Bases:
allennlp.predictors.predictor.Predictor
Predictor for the
target_extraction.allen.models.target_tagger.TargetTagger
model. This predictor is very much based on thefrom allennlp.predictors.sentence.SentenceTaggerPredictor
The main difference: 1. The option to use either the tokenizer that is in the constructor of theclass or to provide the tokens within the JSON that is to be processed thus allowing the flexiability of using your own custom tokenizer.