klue-roberta-small-ner-identified
This model is a fine-tuned version of vitus9988/klue-roberta-small-ner-identified on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0082
- Precision: 0.9930
- Recall: 0.9988
- F1: 0.9959
- Accuracy: 0.9988
Model description
개인정보 비식별을 위해 아래 항목에 대한 개체명 인식을 제공합니다.
- 사람이름 [PS]
- 주소 (구 주소 및 도로명 주소) [AD]
- 카드번호 [CN]
- 계좌번호 [BN]
- 운전면허번호 [DN]
- 주민등록번호 [RN]
- 여권번호 [PN]
- 전화번호 [PH]
- 이메일 주소 [EM]
- 날짜 [DT]
License
This model is a fine-tuned derivative of
klue/roberta-small. Since the base model does not currently specify a license, no license is specified for this derivative model at this time.
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 5e-05
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 61 | 0.0128 | 0.9871 | 0.9929 | 0.9900 | 0.9979 |
| No log | 2.0 | 122 | 0.0098 | 0.9895 | 0.9976 | 0.9935 | 0.9987 |
| No log | 3.0 | 183 | 0.0082 | 0.9930 | 0.9988 | 0.9959 | 0.9988 |
Framework versions
- Transformers 4.40.2
- Pytorch 2.3.0+cu118
- Datasets 2.19.1
- Tokenizers 0.19.1
Use
from transformers import AutoTokenizer, AutoModelForTokenClassification
from transformers import pipeline
tokenizer = AutoTokenizer.from_pretrained("vitus9988/klue-roberta-small-ner-identified")
model = AutoModelForTokenClassification.from_pretrained("vitus9988/klue-roberta-small-ner-identified")
nlp = pipeline("ner", model=model, tokenizer=tokenizer, aggregation_strategy="simple")
example = """
저는 김철수입니다. 집은 서울특별시 강남대로이고 전화번호는 010-1234-5678, 주민등록번호는 123456-1234567입니다. 메일주소는 hugging@face.com입니다. 저는 10월 25일에 출국할 예정입니다.
"""
ner_results = nlp(example)
for i in ner_results:
print(i)
#{'entity_group': 'PS', 'score': 0.9617835, 'word': '김철수', 'start': 3, 'end': 6}
#{'entity_group': 'AD', 'score': 0.9839702, 'word': '서울특별시 강남대로', 'start': 14, 'end': 24}
#{'entity_group': 'PH', 'score': 0.9906756, 'word': '010 - 1234 - 5678', 'start': 33, 'end': 46}
#{'entity_group': 'RN', 'score': 0.9904553, 'word': '123456 - 1234567', 'start': 56, 'end': 70}
#{'entity_group': 'EM', 'score': 0.99022245, 'word': 'hugging @ face. com', 'start': 81, 'end': 97}
#{'entity_group': 'DT', 'score': 0.985629, 'word': '10월 25일', 'start': 105, 'end': 112}