KhoiBui/tiktok-text-safety-classifier

TikTok Content Safety - Text Classifier

This model classifies TikTok video text content (captions, comments) as safe or harmful.

Usage

from transformers import AutoTokenizer, AutoModelForSequenceClassification

tokenizer = AutoTokenizer.from_pretrained("KhoiBui/tiktok-text-safety-classifier")
model = AutoModelForSequenceClassification.from_pretrained("KhoiBui/tiktok-text-safety-classifier")

text = "your text here"
inputs = tokenizer(text, return_tensors="pt", truncation=True, max_length=512)
outputs = model(**inputs)
prediction = outputs.logits.argmax(-1).item()  # 0=safe, 1=harmful

Model Details

• Base Model: XLM-RoBERTa / CafeBERT
• Task: Binary classification (safe/harmful)
• Training Data: TikTok video metadata (Vietnamese + English)