tartunlp Goto Github PK

alpaca-est

Alpaca-style instruction-following dataset for Estonian

bert-ner-api

BERT-based named entity recognition API

bleu2vec

Modified BLEU score. Uses word2vec similarity instead of one to one similarity.

catlm

An implementation of an RNN language model - including training, sampling and scoring texts; plus categories.

coupled-seq2seq

Implementation of coupled sequence-to-sequence model training, based on HuggingFace.

deepvoice3_pytorch

Deep Voice 3 adaptation for Estonian text-to-speech.

domain_clusters

epub-api

E-book to audiobook api.

epub-worker

Processing of e-book to audio.

estgec

Grammatical error correction resources for Estonian and other languages

estner

estner_new

New Estonian NER dataset

estnltk-bert-tagger-webservice

estnltk-stanza-tagger-webservice

Webservices for EstNLTK's StanzaSyntaxTagger and StanzaSyntaxEnsembleTagger

fairseq

Facebook AI Research Sequence-to-Sequence Toolkit written in Python.

finno-ugric-data

Finno-Ugric data

firefox-translations-training

Training pipelines for Firefox Translations neural machine translation models

grammar-api

grammar-worker

inherent-domains-wmt21

This is the official code release for the "Translation Transformers Rediscover Inherent Data Domains", WMT21

kama

Models, scripts and reports for multilingual multi-domain neural machine translation, developed at TartuNLP / University of Tartu

keelesugulased

Linguistic data for Finno-Ugric languages close to Estonian

kiirkirjutaja4livesubs

livesubs

llammas

Adapting Llama-2 to Estonian

m2m-100-finetune

Scripts for finetuning m2m-100 models

memoq-neurotolge-plugin

Plugin using Neurotõlge Machine Translation in MemoQ

memsource-neurotolge-integration

External integration between the Memsource translators' framework and Neurotõlge machine translation service.

meta-share

Public repository of the META-SHARE software

mix-and-match-nmt

Combining encoders and decoders of different pre-trained NMT models to create mix-and-match models.