sentence2tags

Given a sentence predict lemma, morphological tags, pos-tag for each word.

This repo is a modification of the COMBO architecture.

Prerequisites

• Python 3.6
• Clone the repository and install the dependencies

git clone https://github.com/MartinXPN/sentence2tags.git
cd sentence2tags
pip install .

Get the data (from https://github.com/UniversalDependencies)

# Armenian data
wget https://raw.githubusercontent.com/UniversalDependencies/UD_Armenian-ArmTDP/master/hy_armtdp-ud-train.conllu -P datasets
wget https://raw.githubusercontent.com/UniversalDependencies/UD_Armenian-ArmTDP/master/hy_armtdp-ud-dev.conllu -P datasets
wget https://raw.githubusercontent.com/UniversalDependencies/UD_Armenian-ArmTDP/master/hy_armtdp-ud-test.conllu -P datasets


# Russian data
wget https://raw.githubusercontent.com/UniversalDependencies/UD_Russian-SynTagRus/master/ru_syntagrus-ud-train.conllu -P datasets
wget https://raw.githubusercontent.com/UniversalDependencies/UD_Russian-SynTagRus/master/ru_syntagrus-ud-dev.conllu -P datasets
wget https://raw.githubusercontent.com/UniversalDependencies/UD_Russian-SynTagRus/master/ru_syntagrus-ud-test.conllu -P datasets

Train the model

PYTHONHASHSEED=0 python -m sentence2tags.main --mode autotrain 
        --train datasets/ru_syntagrus-ud-train.conllu --valid datasets/ru_syntagrus-ud-dev.conllu 
        --embed datasets/ru.vec --model ru.pkl --force_trees 
        --targets lemma upostag feats --loss_weights 0.2 0.2 0.8

Make predictions

PYTHONHASHSEED=0 python -m sentence2tags.main --mode predict 
        --test test_data.conllu --pred output_path.conllu --model ru.pkl