Sms Spam Classifier

🚀 SMS Spam Classifier

This model is a multilingual BERT-based model fine-tuned for SMS spam detection. It can classify SMS messages as ham (non-spam) or spam. It is trained based on the bert-base-multilingual-cased model from the Hugging Face Transformers library.

🚀 Quick Start

This model can be used directly through the Hugging Face Transformers library:

from transformers import AutoTokenizer, AutoModelForSequenceClassification

# Load the model and tokenizer
tokenizer = AutoTokenizer.from_pretrained("blockenters/sms-spam-classifier")
model = AutoModelForSequenceClassification.from_pretrained("blockenters/sms-spam-classifier")

# Input sample
text = "축하합니다! 무료 발리 여행 티켓을 받으셨습니다. WIN이라고 회신하세요."

# Tokenize and predict
inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=128)
outputs = model(**inputs)
predictions = outputs.logits.argmax(dim=-1)

# Decode the prediction result
label_map = {0: "ham", 1: "spam"}
print(f"Prediction result: {label_map[predictions.item()]}")

✨ Features

• The data used for training was created by directly processing Korean SMS. If you are interested in the dataset, please contact us.
• This model can classify SMS messages as ham or spam effectively.

📚 Documentation

📦 Model Details

📦 Dataset Information

The dataset used for training and evaluation contains SMS messages labeled as ham (non-spam) or spam. After pre - processing, the data was split as follows:

• Training Data: 80%
• Validation Data: 20%

📦 Training Settings

• Learning Rate: 2e-5
• Batch Size: 8 (per device)
• Epochs: 1
• Evaluation Strategy: Per epoch
• Tokenizer: bert-base-multilingual-cased

This model was efficiently fine - tuned using Hugging Face's Trainer API.

📄 License

This model is licensed under the apache-2.0 license.