V3.0.0.5 - bump version

This is a bugfix release

V3.0.0.4 - fix gensim version

This is a bugfix release.

V3.0.0.3 - fix crfutils not included in sdist

This is a bugfix release.

V3.0.0.2 - fix relative import issue

This is a bugfix release.

• Fixes #31

V3.0.0.1 - fix requirements

Fix requirements and setup.py.

• This is a bugfix release (system code and performance identical to 3.0.0.0).

V3.0.0.0 - single model MTL sentencer, POS tagger and morph tagger

This version is a major architectural upgrade to V2.X - the neural POS tagger, morphological tagger and sentence splitter have been replaced with a single transformer based model (work by @nitinvwaran) trained on all three tasks using Multitask Learning (MTL). This lowers the memory footprint of the system, substantially speeds up start up and run-time, and allows the system to run on GPU with a more modest GPU RAM (~4 GB). Accuracy is on par with or slightly better than the previous version.

V2.0.0.2 - requirements fix

• More specific library versions requirements, otherwise identical to 2.0.0.1.
• Fixes #19

V2.0.0.1 - Pure python neural components

• Totally removed Java dependencies
• New neural components for parser, morphology and segmentation

V1.0.0.3 - python 3.8 support

Fixes #9 issue with timing module for python 3.8 support

V1.0.0.2 - runnable as a module, installable from PyPI

HebPipe is now runnable as a module and installable from PyPI. Java 64 bit is required, models and Python dependencies can be downloaded automatically.

• Simplified dependencies
• Update segmentation to use XGBoost (model uses V0.81)
• Set up PyPI support, install and run with:
  • pip install hebpipe
  • python -m hebpipe