jisukim8873's picture
Update README.md
b299a11
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, TakSung Heo, and Minsu Jeong
  • Model Type: Text-generation
  • Language: Korean
  • License: CC-BY-4.0

Dataset

BLEU Score

  • 25.35

Uses

This model can be used for convert speech style

  • formal: 문어체
  • informal: ꡬ어체
  • android: μ•ˆλ“œλ‘œμ΄λ“œ
  • azae: μ•„μž¬
  • chat: μ±„νŒ…
  • choding: μ΄ˆλ“±ν•™μƒ
  • emoticon: 이λͺ¨ν‹°μ½˜
  • enfp: enfp
  • gentle: 신사
  • halbae: 할아버지
  • halmae: ν• λ¨Έλ‹ˆ
  • joongding: 쀑학생
  • king: μ™•
  • naruto: λ‚˜λ£¨ν† 
  • seonbi: μ„ λΉ„
  • sosim: μ†Œμ‹¬ν•œ
  • translator: λ²ˆμ—­κΈ°

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", "ꡬ어체", "μ•ˆλ“œλ‘œμ΄λ“œ", "μ•„μž¬", "μ±„νŒ…", "μ΄ˆλ“±ν•™μƒ", "이λͺ¨ν‹°μ½˜", "enfp", "신사", "할아버지", "ν• λ¨Έλ‹ˆ", "쀑학생", "μ™•", "λ‚˜λ£¨ν† ", "μ„ λΉ„", "μ†Œμ‹¬ν•œ", "λ²ˆμ—­κΈ°"]

for style in styles:
    text = f"{style} ν˜•μ‹μœΌλ‘œ λ³€ν™˜:μ˜€λŠ˜μ€ λ‹­λ³ΆμŒνƒ•μ„ λ¨Ήμ—ˆλ‹€. λ§›μžˆμ—ˆλ‹€."
    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