README.md

---
language:
  - ko
tags:
- generated_from_keras_callback
model-index:
- name: RoBERTa-large-Detection-P2G
  results: []
---

# RoBERTa-large-Detection-P2G


이 모델은 klue/roberta-large을 국립 국어원 신문 말뭉치 5만개의 문장을 2021을 g2pK로 훈련시켜 G2P된 데이터를 탐지합니다.<br>
git : https://github.com/taemin6697<br>

## Usage
```python
from transformers import AutoTokenizer, RobertaForSequenceClassification
import torch
import numpy as np

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
model_dir = "kfkas/RoBERTa-large-Detection-G2P"
tokenizer = AutoTokenizer.from_pretrained('klue/roberta-large')
model = RobertaForSequenceClassification.from_pretrained(model_dir).to(device)

text = "월드커 파나은행 대표티메 행우늬 이달러 이영영장 선물"
with torch.no_grad():
    x = tokenizer(text, padding='max_length', truncation=True, return_tensors='pt', max_length=128)
    y_pred = model(x["input_ids"].to(device))
    logits = y_pred.logits
    y_pred = logits.detach().cpu().numpy()
    y = np.argmax(y_pred)
    print(y)
    #1
```


## Intended uses & limitations

More information needed

## Training and evaluation data

More information needed

## Training procedure

### Training hyperparameters

The following hyperparameters were used during training:
- optimizer: None
- training_precision: float16
### Training results
### Framework versions
- Transformers 4.22.1
- TensorFlow 2.10.0
- Datasets 2.5.1
- Tokenizers 0.12.1