Part-of-speech tagger and lemmatizer for Latin.
- POS tagging
- Postprocessing
- Evaluation
- Errors
- Output and evaluation script
- Facilitate engagement
- Lint and clean up
- Preprocessing
- Start/end sentence boundaries
- Re-train
- Increase token embedding, character embedding and hidden sizes
- Fix batch size to 8
- Lemmatization
- Preprocessing
- External unlabelled data
- External labelled data
- Data augmentation methods
0.957
0.905