Model Card for Model ID

μ˜μ–΄-ν•œκ΅­μ–΄ λ²ˆμ—­ λͺ¨λΈμž…λ‹ˆλ‹€.

Model Description

paust/pko-t5-base λͺ¨λΈμ„ 기반으둜 μ˜μ–΄-ν•œκ΅­μ–΄ λ²ˆμ—­μ„ λ―Έμ„Έμ‘°μ •ν•œ λ²ˆμ—­ λͺ¨λΈμž…λ‹ˆλ‹€. μ˜μ–΄->ν•œκ΅­μ–΄, ν•œκ΅­μ–΄->μ˜μ–΄ μ–‘λ°©ν–₯ λ²ˆμ—­μ„ μ§€μ›ν•˜λ©°, μ˜μ–΄->ν•œκ΅­μ–΄ λ²ˆμ—­ μ‹œ λ†’μž„λ§λ„ μ„€μ •ν•  수 μžˆμŠ΅λ‹ˆλ‹€.

  • Developed by: [BlueAI]
  • Model type: [t5.1.1.base]
  • Language(s) (NLP): [Korean]
  • License: [MIT]
  • Finetuned from model [optional]: [paust/pko-t5-base]

Uses

How to Get Started with the Model

Use the code below to get started with the model.

from transformers import pipeline, T5TokenizerFast

tokenizer_name = "paust/pko-t5-base"
tokenizer = T5TokenizerFast.from_pretrained(tokenizer_name)
model_path = 'Darong/BlueT'
translator = pipeline("translation", model=model_path, tokenizer=tokenizer, max_length=255)
# μ˜μ–΄ -> ν•œκ΅­μ–΄
prefix = "E2K: "
source = "This model is an English-Korean translation model."
target = translator(prefix + source)
print(target[0]['translation_text'])

# ν•œκ΅­μ–΄->μ˜μ–΄
prefix = "K2E: "
source = "이 λͺ¨λΈμ€ μ˜μ–΄-ν•œκ΅­μ–΄ λ²ˆμ—­ λͺ¨λΈμž…λ‹ˆλ‹€."
target = translator(prefix + source)
print(target[0]['translation_text'])

Training Details

Training Data

이 λͺ¨λΈμ€ AI Hub 및 자체 κ΅¬μΆ•ν•œ λ°μ΄ν„°λ‘œ ν•™μŠ΅λ˜μ—ˆμŠ΅λ‹ˆλ‹€. μ˜μ–΄->ν•œκ΅­μ–΄ ν•™μŠ΅ 데이터 μˆ˜λŠ” 1800만 이상, ν•œκ΅­μ–΄->μ˜μ–΄ ν•™μŠ΅ 데이터 μˆ˜λŠ” 1200만 μ΄μƒμ˜ λ¬Έμž₯으둜 κ΅¬μΆ•λ˜μ—ˆμŠ΅λ‹ˆλ‹€.

Downloads last month
185
Safetensors
Model size
276M params
Tensor type
F32
Β·
Inference Providers NEW
This model isn't deployed by any Inference Provider. πŸ™‹ Ask for provider support

Model tree for Darong/BlueT

Base model

paust/pko-t5-base
Finetuned
(12)
this model