pyabsa.tasks.ABSAInstruction.model¶
Classes¶
Module Contents¶
- class pyabsa.tasks.ABSAInstruction.model.T5Generator(checkpoint)¶
- tokenizer¶
- model¶
- data_collator¶
- device¶
- tokenize_function_inputs(sample)¶
Udf to tokenize the input dataset.
- train(tokenized_datasets, **kwargs)¶
Train the generative model.
- predict(text, **kwargs)¶
Predict the output from the model.
- get_labels(tokenized_dataset, trained_model_path=None, predictor=None, batch_size=4, sample_set='train')¶
Get the predictions from the trained model.
- get_aspect_metrics(true_aspects, pred_aspects)¶
- get_classic_metrics(y_true, y_pred)¶
- class pyabsa.tasks.ABSAInstruction.model.T5Classifier(model_checkpoint)¶
- tokenizer¶
- model¶
- data_collator¶
- tokenize_function_inputs(sample)¶
Udf to tokenize the input dataset.
- train(tokenized_datasets, **kwargs)¶
Train the generative model.
- get_labels(tokenized_dataset, predictor=None, batch_size=4, sample_set='train')¶
Get the predictions from the trained model.
- get_metrics(y_true, y_pred)¶
- class pyabsa.tasks.ABSAInstruction.model.ABSAGenerator(checkpoint)¶
Bases:
T5Generator