v2.1.1

reldi-tokeniser 1.0.3 added as dependency, in which a bug in abbreviation loading has been resolved.

v2.1

• Added new models for all languages
• Added new "web" processing type
• Fixed sentence splitting in the tokenizers

v2.0

• Added new models for standard Slovenian
• Added new inflectional lexicon for Slovenian
• Adapted tests to new model outputs
• Modified lexicon to store underscores instead of empty strings
• Other changes

v1.2.0

• Added SRL parsing to Slovenian language
• Fixed training for lemmatizer and pos tagger
• Added toy tests for all trainings
• Other smaller fixes

v1.1.1

• Updated external package version requirements. Mainly due to updates in Slovenian obeliks tokenizer

v1.1.0

• Added tokenizer pretag option for both obeliks and reldi-tokeniser (via pos_lemma_pretag)
• Updated Slovene inflectional lexicon and moved from lemmatizer model to morphosyntactic annotation model
• Added upos and ufeats control to Slovene inflectional lexicon
• Other smaller fixes

v1.0.2

• fixed issue where the parser produced non-CONLLU-compliant extension labels with underscores (e.g. cc_preconj) instead of colon-separated labels (e.g. cc:preconj)
• during lemmatization, if a token consists of a character that is not present in the seq2seq vocabulary, lemma will now be identical to the token
• added PUNCT control
• fixed MISC collumn bug for NER
• punct in Bulgarian UPOS was renamed to Z