law-ai/semantic-segmentation

Semantic Segmentation of Legal texts that labels sentences with one of 7 rhetorical roles.

semantic-segmentation's Issues

Index out-of-bounds error in prepare_data.py

I encountered an issue in the prepare_data.py where an index out-of-bounds error occurs. After investigating, I found that the line causing the error is:
f = (f + 1) % 5
To fix this issue and avoid the out-of-bounds error, I suggest updating the line to use a variable docs_per_fold instead of the hardcoded value 5:
f = (f + 1) % docs_per_fold

Issue when using the trained model for inference

There seems to be an issue when I'm using the trained model on unannotated data for inference.
Also the link provided for the Sen2vec for using the pre-trained model does not work.
Could you please help me with this