nsfw classifier

Introduction
The NSFW Classifier is an AI-powered image classification model designed to detect and filter NSFW content such as pornography, hentai, and sexually suggestive images. It is aimed at maintaining the safety and legal compliance of platforms that allow user-uploaded images.

Architecture
The model leverages ViT (Vision Transformer) architecture and is implemented using PyTorch. It operates with an accuracy metric of 92% and utilizes the deepghs/nsfw_detect dataset for training. It is integrated with Hugging Face's Transformers library and can be monitored via TensorBoard.

Training
The model was trained on the deepghs/nsfw_detect dataset, focusing on accurate classification of NSFW images. It achieves a high accuracy of 92%, making it a reliable tool for detecting explicit content.

Guide: Running Locally

1. Setup Environment: Install PyTorch and Transformers library.
2. Clone Repository: Clone the NSFW Classifier repository from Hugging Face.
3. Download Model: Load the NSFW Classifier model weights.
4. Inference: Run inference using the provided image classification scripts.
5. Cloud GPUs: For enhanced performance, consider using cloud GPU services like AWS EC2, Google Cloud Platform, or Azure.

License
The NSFW Classifier is distributed under the CC BY-NC-ND 4.0 license. This license permits use with attribution, prohibits commercial use, and does not allow derivative works.