langtest.modelhandler.transformers_modelhandler.PretrainedModelForTextClassification#

class PretrainedModelForTextClassification(model: Pipeline)#

Bases: ModelAPI

Transformers pretrained model for text classification tasks

model#

Loaded Text Classification pipeline for predictions.

Type:

transformers.pipeline.Pipeline

__init__(model: Pipeline)#

Constructor method

Parameters:

model (transformers.pipeline.Pipeline) – Pretrained HuggingFace NER pipeline for predictions.

Methods

__init__(model)	Constructor method
load_model(path)	Load and return text classification transformers pipeline
predict(text[, return_all_scores, ...])	Perform predictions on the input text.
predict_raw(text[, truncation_strategy])	Perform predictions on the input text.

Attributes

labels	Return classification labels of pipeline model.
model_registry

property labels: List[str]#

Return classification labels of pipeline model.

classmethod load_model(path: str) → Pipeline#

Load and return text classification transformers pipeline

predict(text: str, return_all_scores: bool = False, truncation_strategy: str = 'longest_first', *args, **kwargs) → SequenceClassificationOutput#

Perform predictions on the input text.

Parameters:

• text (str) – Input text to perform NER on.

• return_all_scores (bool) – Option to group entities.

• truncation_strategy (str) – strategy to use to truncate too long sequences

• kwargs – Additional keyword arguments.

Returns:

text classification from the input text.

Return type:

SequenceClassificationOutput

predict_raw(text: str, truncation_strategy: str = 'longest_first') → List[str]#

Perform predictions on the input text.

Parameters:

• text (str) – Input text to perform NER on.

• truncation_strategy (str) – strategy to use to truncate too long sequences

Returns:

Predictions as a list of strings.

Return type:

List[str]