sappho192's picture
Update README.md
80ee466 verified
|
raw
history blame
3.82 kB
metadata
license: mit
language:
  - ja
  - ko
pipeline_tag: translation

Japanese to Korean translator

Japanese to Korean translator model based on EncoderDecoderModel(bert-japanese+kogpt2)

Usage

Demo

Please visit https://huggingface.co/spaces/sappho192/aihub-ja-ko-translator-demo

Dependencies (PyPI)

  • torch
  • transformers
  • fugashi
  • unidic-lite

Inference

from transformers import(
    EncoderDecoderModel,
    PreTrainedTokenizerFast,
    BertJapaneseTokenizer,
)

import torch

encoder_model_name = "cl-tohoku/bert-base-japanese-v2"
decoder_model_name = "skt/kogpt2-base-v2"

src_tokenizer = BertJapaneseTokenizer.from_pretrained(encoder_model_name)
trg_tokenizer = PreTrainedTokenizerFast.from_pretrained(decoder_model_name)

model = EncoderDecoderModel.from_pretrained("sappho192/aihub-ja-ko-translator")

text = "εˆγ‚γΎγ—γ¦γ€‚γ‚ˆγ‚γ—γγŠι‘˜γ„γ—γΎγ™γ€‚"

def translate(text_src):
    embeddings = src_tokenizer(text_src, return_attention_mask=False, return_token_type_ids=False, return_tensors='pt')
    embeddings = {k: v for k, v in embeddings.items()}
    output = model.generate(**embeddings, max_length=500)[0, 1:-1]
    text_trg = trg_tokenizer.decode(output.cpu())
    return text_trg

print(translate(text))

Dataset

This model used datasets from 'The Open AI Dataset Project (AI-Hub, South Korea)'.
All data information can be accessed through 'AI-Hub (aihub.or.kr)'.
(In order for a corporation, organization, or individual located outside of Korea to use AI data, etc., a separate agreement is required with the performing organization and the Korea National Information Society agency(NIA). In order to export AI data, etc. outside the country, a separate agreement is required with the performing organization and the NIA. Link)

이 λͺ¨λΈμ€ κ³Όν•™κΈ°μˆ μ •λ³΄ν†΅μ‹ λΆ€μ˜ μž¬μ›μœΌλ‘œ ν•œκ΅­μ§€λŠ₯μ •λ³΄μ‚¬νšŒμ§„ν₯μ›μ˜ 지원을 λ°›μ•„ κ΅¬μΆ•λœ 데이터셋을 ν™œμš©ν•˜μ—¬ μˆ˜ν–‰λœ μ—°κ΅¬μž…λ‹ˆλ‹€.
λ³Έ λͺ¨λΈμ— ν™œμš©λœ λ°μ΄ν„°λŠ” AI ν—ˆλΈŒ(aihub.or.kr)μ—μ„œ λ‹€μš΄λ‘œλ“œ λ°›μœΌμ‹€ 수 μžˆμŠ΅λ‹ˆλ‹€.
(ꡭ외에 μ†Œμž¬ν•˜λŠ” 법인, 단체 λ˜λŠ” 개인이 AI데이터 등을 μ΄μš©ν•˜κΈ° μœ„ν•΄μ„œλŠ” μˆ˜ν–‰κΈ°κ΄€ λ“± 및 ν•œκ΅­μ§€λŠ₯μ •λ³΄μ‚¬νšŒμ§„ν₯원과 λ³„λ„λ‘œ ν•©μ˜κ°€ ν•„μš”ν•©λ‹ˆλ‹€.
λ³Έ AI데이터 λ“±μ˜ κ΅­μ™Έ λ°˜μΆœμ„ μœ„ν•΄μ„œλŠ” μˆ˜ν–‰κΈ°κ΄€ λ“± 및 ν•œκ΅­μ§€λŠ₯μ •λ³΄μ‚¬νšŒμ§„ν₯원과 λ³„λ„λ‘œ ν•©μ˜κ°€ ν•„μš”ν•©λ‹ˆλ‹€. [좜처])

Dataset list

The dataset used to train the model is merged following sub-datasets:

    1. μΌμƒμƒν™œ 및 ꡬ어체 ν•œ-쀑, ν•œ-일 λ²ˆμ—­ 병렬 λ§λ­‰μΉ˜ 데이터 [Link]
    1. ν•œκ΅­μ–΄-λ‹€κ΅­μ–΄(μ˜μ–΄ μ œμ™Έ) λ²ˆμ—­ λ§λ­‰μΉ˜(κΈ°μˆ κ³Όν•™) [Link]
    1. ν•œκ΅­μ–΄-λ‹€κ΅­μ–΄ λ²ˆμ—­ λ§λ­‰μΉ˜(κΈ°μ΄ˆκ³Όν•™) [Link]
    1. ν•œκ΅­μ–΄-λ‹€κ΅­μ–΄ λ²ˆμ—­ λ§λ­‰μΉ˜ (인문학) [Link]
  • ν•œκ΅­μ–΄-일본어 λ²ˆμ—­ λ§λ­‰μΉ˜ [Link]

To reproduce the the merged dataset, you can use the code in below link:
https://github.com/sappho192/aihub-translation-dataset