This is a demo of the project to create a sociolinguistic classifier for the Russian language. At this stage, the possibility of recognizing the labels "work" and "housing" is embodied, as well as Named Entity Recognition task. Feel free to try it out!
We use tiny Rubert with classification head and SHAP for prediction interpretation
NER-model is Rubert-collection3 — fine-tuned version of sberbank-ai/ruBert-base
