- Fine-tunning "KLUE/roberta-large" model For CER(Company Entity Recognition) With Custom Dataset
- Custom Datasets are composed of news data
label_list = ['O',"B-PER","I-PER","B-ORG","I-ORG","B-COM","I-COM","B-LOC","I-LOC","B-DAT","I-DAT","B-TIM","I-TIM","B-QNT","I-QNT"]
refer_list = ['0','1','2','3','4','5','6','7','8','9','10','11','12','13','14']
- EX: "B-PER" : 1 , "B-COM" : 5