metadata
language:
- ko
tags:
- korean
- token-classification
- pos
- dependency-parsing
base_model: KoichiYasuoka/roberta-base-korean-hanja
license: cc-by-sa-4.0
pipeline_tag: token-classification
widget:
- text: 홍시 맛이 나서 홍시라 생각한다.
- text: 紅柹 맛이 나서 紅柹라 生覺한다.
roberta-base-korean-morph-upos
Model Description
This is a RoBERTa model pre-trained on Korean texts for POS-tagging and dependency-parsing, derived from roberta-base-korean-hanja and morphUD-korean. Every morpheme (형태소) is tagged by UPOS(Universal Part-Of-Speech).
How to Use
from transformers import AutoTokenizer,AutoModelForTokenClassification,TokenClassificationPipeline
tokenizer=AutoTokenizer.from_pretrained("KoichiYasuoka/roberta-base-korean-morph-upos")
model=AutoModelForTokenClassification.from_pretrained("KoichiYasuoka/roberta-base-korean-morph-upos")
pipeline=TokenClassificationPipeline(tokenizer=tokenizer,model=model,aggregation_strategy="simple")
nlp=lambda x:[(x[t["start"]:t["end"]],t["entity_group"]) for t in pipeline(x)]
print(nlp("홍시 맛이 나서 홍시라 생각한다."))
or
import esupar
nlp=esupar.load("KoichiYasuoka/roberta-base-korean-morph-upos")
print(nlp("홍시 맛이 나서 홍시라 생각한다."))
See Also
esupar: Tokenizer POS-tagger and Dependency-parser with BERT/RoBERTa/DeBERTa models