Release History

Note that prior to version 1.0.0, the Stanza library was named as “StanfordNLP”. To install historical versions prior to to v1.0.0, you’ll need to run pip install stanfordnlp.

VersionDate          Notes
1.4.02022‑04‑22Transformers added to NER and Constituency Parser. (full release log)
1.3.02021‑10‑05Multilingual Pipeline, Constituency Parser. (full release log)
1.2.32021‑08‑09Additional NER models for AF, IT. A couple minor bugfixes. Additional NER finetuning support. (full release log)
1.2.22021-07-15Fix regression in NER results for non-VI languages, fix a couple other minor bugs (full release log)
1.2.12021-06-17Additional NER models for BG, HU, FI, VI. bulk_process mechanism which works much faster in certain settings. All models updated to UD 2.8 data. (full release log)
1.2.02021-01-29Rewriting of the training scripts to python for greater ease of use and system compatibility. Multiple fixes to tokenization models for common punctuation errors and typos. All models updated to UD 2.7 data. (full release log)
1.1.12020-08-13This release features support for extending the capability of the Stanza pipeline with customized processors, a new sentiment analysis tool for English/German/Chinese, improvements to the CoreNLPClient functionality (including compatibility with CoreNLP 4.1.0), new models for a few languages (including Thai, which is supported for the first time in Stanza), new biomedical and clinical English packages, alternative servers for downloading resource files, and various improvements and bugfixes (full release log).
1.0.12020-04-27This is a maintenance release of Stanza. It features new support for jieba as Chinese tokenizer, faster lemmatizer implementation, improved compatibility with CoreNLP v4.0.0, and several bugfixes including correct character offsets in NER output and correct Vietnamese tokenization outputs (full release log).
1.0.02020-03-17This release introduces new multi-lingual named entity recognition (NER) support for 8 languages, expanded UD pipeline coverage of 66 languages, improved download and pipeline interfaces, improved document object interfaces, Anaconda installation support, improved neural lemmatizer, spaCy tokenization integration, and various other enhancements and bugfixes (full release log).
0.2.02019-05-16This release introduces substantially reduced model size, substantial lemmatizer speed up and more options for customizing server start up and requests (full release log).
0.1.22019-02-26This release introduces support for pretokenized text, speed ups in the POS/Feats tagger and various bug fixes (full release log).
0.1.02019-01-29Initial release of StanfordNLP (full release log).