README.md

metadata

license: cc-by-4.0
language:
  - ko
tags:
  - generation

Model Details

• Model Description: Speech style converter model based on gogamza/kobart-base-v2
• Developed by: Juhwan, Lee and Jisu, Kim
• Model Type: Text-generation
• Language: Korean
• License: CC-BY-4.0

Dataset

• korean SmileStyle Dataset
• Randomly split train/valid dataset (9:1)

BLEU Score

• 25.35

Uses

This model can be used for convert speech style

• formal: 문어체
• informal: 구어체
• android: 안드로이드
• azae: 아재
• chat: 채팅
• choding: 초등학생
• emoticon: 이모티콘
• enfp: enfp
• gentle: 신사
• halbae: 할아버지
• halmae: 할머니
• joongding: 중학생
• king: 왕
• naruto: 나루토
• seonbi: 선비
• sosim: 소심한
• translator: 번역기

from transformers import pipeline

model = "KoJLabs/bart-speech-style-converter"
tokenizer = AutoTokenizer.from_pretrained(model)

nlg_pipeline = pipeline('text2text-generation',model=model, tokenizer=tokenizer)
styles = ["문어체", "구어체", "안드로이드", "아재", "채팅", "초등학생", "이모티콘", "enfp", "신사", "할아버지", "할머니", "중학생", "왕", "나루토", "선비", "소심한", "번역기"]

for style in styles:
    text = f"{style} 형식으로 변환:오늘은 닭볶음탕을 먹었다. 맛있었다."
    out = nlg_pipeline(text, max_length=100)
    print(style, out[0]['generated_text'])

Model Source

https://github.com/KoJLabs/speech-style/tree/main

Speech style conversion package

You can exercise korean speech style conversion task with python package KoTAN