theSOL1
/

kogrammar-base

Text2Text Generation Transformers PyTorch Safetensors Korean bart grammar Inference Endpoints

Model card Files Files and versions Community

Edit model card

kogrammar-base

Dataset: 국립국어원 맞춤법 교정 말뭉치

Backbone Model: kobart-base-v2
GitHub Repo: SOL1archive/KoGrammar

Train Method

전체 데이터셋 중 약 45%를 학습데이터로 활용하여 학습함.

Metric

BLEU-2	ROUGE-2 F1
77.8 %	55.0 %

Downloads last month: 276

Safetensors

Model size

124M params

Tensor type

F32

·