{ "version": "1.0", "truncation": null, "padding": null, "added_tokens": [ { "id": 44, "content": "", "single_word": false, "lstrip": false, "rstrip": false, "normalized": false, "special": true } ], "normalizer": { "type": "Sequence", "normalizers": [ { "type": "Lowercase" }, { "type": "Replace", "pattern": { "Regex": "[^\u0447 \u0434\u044c\u044f\u0439\u0441\u0443\u0430\u043e20\u044a\u044b\u043d\u043b\u0436\u0445\u0448\u0437c_\u043fm\u044e\u044d\u04404\u0449o\u2013\u0444q\u0438\u0446\u043a1\u0435\u0431\\-\u0442\u0432\u0433\u043c]" }, "content": "" }, { "type": "Strip", "strip_left": true, "strip_right": true }, { "type": "Replace", "pattern": { "Regex": "(?=.)|(?": 44 } } }