This PR adds an optional parameter to the SequenceLabeler.predict method in the SequenceLabeler class. The new parameter, spans, is a list of dictionaries where each dictionary contains the start and end indices of the predefined spans. This feature allows for predicting the labels of predefined spans in scenarios where the segmentation of entities is already known.
Summary of Changes
Modified the SequenceLabeler class in finetune/target_models/sequence_labeling.py to include the spans parameter in the predict method.
Updated the predict method to handle the case when the spans parameter is provided. The method now only makes predictions for the provided spans.
Updated the _predict and _predict_decode methods in finetune/target_models/sequence_labeling.py to handle the case when the spans parameter is provided.
Updated the sequence_labeler method in finetune/nn/target_blocks.py to handle the case when the spans parameter is provided.
Fixes #1.
To checkout this PR branch, run the following command in your terminal:
git checkout sweep/add-predict-spans-param
To get Sweep to edit this pull request, leave a comment below or in the code. Leaving a comment in the code will only modify the file but commenting below can change the entire PR.
Description
This PR adds an optional parameter to the
SequenceLabeler.predictmethod in theSequenceLabelerclass. The new parameter,spans, is a list of dictionaries where each dictionary contains the start and end indices of the predefined spans. This feature allows for predicting the labels of predefined spans in scenarios where the segmentation of entities is already known.Summary of Changes
SequenceLabelerclass infinetune/target_models/sequence_labeling.pyto include thespansparameter in thepredictmethod.predictmethod to handle the case when thespansparameter is provided. The method now only makes predictions for the provided spans._predictand_predict_decodemethods infinetune/target_models/sequence_labeling.pyto handle the case when thespansparameter is provided.sequence_labelermethod infinetune/nn/target_blocks.pyto handle the case when thespansparameter is provided.Fixes #1.
To checkout this PR branch, run the following command in your terminal:
To get Sweep to edit this pull request, leave a comment below or in the code. Leaving a comment in the code will only modify the file but commenting below can change the entire PR.