KoichiYasuoka's picture
bug fix
3e12ff8
metadata
language:
  - ko
tags:
  - korean
  - token-classification
  - pos
  - dependency-parsing
base_model: team-lucid/deberta-v3-base-korean
datasets:
  - universal_dependencies
license: apache-2.0
pipeline_tag: token-classification
widget:
  - text: 홍시 맛이 나서 홍시라 생각한다.
  - text: 紅柹 맛이 나서 紅柹라 生覺한다.

deberta-base-korean-upos

Model Description

This is a DeBERTa(V3) model pre-trained on Korean texts for POS-tagging and dependency-parsing, derived from deberta-v3-base-korean. Every word (어절) is tagged by UPOS(Universal Part-Of-Speech).

How to Use

from transformers import AutoTokenizer,AutoModelForTokenClassification,TokenClassificationPipeline
tokenizer=AutoTokenizer.from_pretrained("KoichiYasuoka/deberta-base-korean-upos")
model=AutoModelForTokenClassification.from_pretrained("KoichiYasuoka/deberta-base-korean-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/deberta-base-korean-upos")
print(nlp("홍시 맛이 나서 홍시라 생각한다."))

See Also

esupar: Tokenizer POS-tagger and Dependency-parser with BERT/RoBERTa/DeBERTa models