medinify icon indicating copy to clipboard operation
medinify copied to clipboard

Python text classification package with a focus on medical text.

Results 12 medinify issues
Sort by recently updated
recently updated
newest added

Hello and thank you for sharing your work! Could you please provide the last checkpoint for the pretrained models? Thank you in advance, Lucia

Update scraper_example.py, dataset_example.py, and classifier_example.py

Update __main__.py to work with new updates

I'm not sure exactly what this entails. Jorge has fine-tuned a BERT, so if you're assigned this issue, he might be a good person to ask.

enhancement
research

Try adding AllenNLP embeddings somewhere in the package (I think in the CNN)

enhancement
research

Ishaan wrote a part of speech vectorizer that was never merged into develop - it needs to be integrated into the new project structure (as a processor)

Ishaan wrote a bi-LSTM over the summer that was never merged into development - it needs to be integrated into the new project structure

Old Models were trained using old structure and are therefore in the wrong format. They need to be re-trained.

Write unit tests for Classifier.py and cnn_classifier.py