README.md

---
license: apache-2.0
tags:
- bert
- kcbert
- unsmile
---

# SJ-Donald/kcbert-large-unsmile

SJ-Donald/kcbert-large-unsmile is pretrained model using follow:

## Models

* [beomi/kcbert-large](https://huggingface.co/beomi/kcbert-large)

## Datasets

* [smilegate-ai/kor_unsmile](smilegate-ai/kor_unsmile)

## How to use

```Python
from transformers import TextClassificationPipeline, BertForSequenceClassification, AutoTokenizer+

model_name = 'SJ-Donald/kcbert-large-unsmile'
model = BertForSequenceClassification.from_pretrained(model_name)

tokenizer = AutoTokenizer.from_pretrained(model_name)

pipe = TextClassificationPipeline(
    model = model,
    tokenizer = tokenizer,
    device = 0,   # cpu: -1, gpu: gpu number
    return_all_scores = True,
    function_to_apply = 'sigmoid'
)

for result in pipe("이래서 여자는 게임을 하면 안된다")[0]:
    print(result)
    
{'label': '여성/가족', 'score': 0.9793611168861389}
{'label': '남성', 'score': 0.006330598145723343}
{'label': '성소수자', 'score': 0.007870828732848167}
{'label': '인종/국적', 'score': 0.010810344479978085}
{'label': '연령', 'score': 0.020540334284305573}
{'label': '지역', 'score': 0.015790466219186783}
{'label': '종교', 'score': 0.014563685283064842}
{'label': '기타 혐오', 'score': 0.04097242280840874}
{'label': '악플/욕설', 'score': 0.019168635830283165}
{'label': 'clean', 'score': 0.014866289682686329}
```