[QUESTION] Language-specific CoreNLPClient with custom properties

[pleyad]

I am trying to create a CoreNLPClient instance based on the models for German, but with the parameter tokenize.whitespace = true. It should perform POS-Tagging for German Text.

In the documentation (https://stanfordnlp.github.io/stanza/client_properties.html#corenlp-server-start-options-pipeline) it is specified that one could either give the constructor a string for the language or a dictionary with the properties, but not both at the same time.

Am I missing something, is there a way to both specify language models and properties?

[AngledLuffa]

It certainly appears there is no way to mix the properties using just the python interface. What you can do is look in the models jar file for the StanfordCoreNLP-german.properties file, extract it using the jar command, then make your own copy with the desired changes. Does that work for you?

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale[bot]]

This issue has been automatically closed due to inactivity.