Flair Mentions Extractor¶
Bases: MentionsExtractor
Flair Mentions extractor
- Requires flair package to be installed *
Parameters:
Name | Type | Description | Default |
---|---|---|---|
extractor_type |
Optional[ExtractorType]
|
Type of extractor to get mentions. One of: - NER: to use Named Entity Recognition model to get the mentions - POS: to get the mentions based on the linguistics |
NER
|
load_models()
¶
Load Flair model to perform the mentions extraction
predict(docs, batch_size=None)
¶
Predict mentions in each document
Parameters:
Name | Type | Description | Default |
---|---|---|---|
docs |
Iterator[Doc]
|
Documents to get mentions of |
required |
batch_size |
Batch size to use |
None
|
Returns:
Type | Description |
---|---|
Spans of the mentions |