metadata
language: ar
tags:
- arabic
- offensive-language
- text-classification
- content-moderation
- arabert
pipeline_tag: text-classification
Arabic Speech Guard 🛡️
Arabic-Speech-Guard is an Arabic text classification model fine-tuned on AraBERT to detect offensive and harmful speech.
Labels
- 0 → Neutral
- 1 → Offensive
Model Details
- Base model:
aubmindlab/bert-base-arabertv2 - Task: Arabic Offensive Speech Detection
- Language: Arabic
Usage
from transformers import AutoTokenizer, AutoModelForSequenceClassification
import torch
model_name = "USERNAME/Arabic-Speech-Guard"
tokenizer = AutoTokenizer.from_pretrained(model_name)
model = AutoModelForSequenceClassification.from_pretrained(model_name)
text = "اكتب النص هنا"
inputs = tokenizer(text, return_tensors="pt", truncation=True, padding=True)
with torch.no_grad():
outputs = model(**inputs)
prediction = torch.argmax(outputs.logits, dim=1).item()
print("Prediction:", "Danger" if prediction == 1 else "Safe")