A

ARPG

Developed by hp-l33
ARPG is an innovative autoregressive image generation framework capable of achieving BERT-style masked modeling through a GPT-like causal architecture.
Downloads 68
Release Time : 3/9/2025

Model Overview

ARPG is an autoregressive image generation framework that combines GPT's causal architecture with BERT's masked modeling techniques to efficiently generate high-quality images.

Model Features

High Performance
Achieves an FID score of 1.94, generating high-quality images
Efficient Generation
Utilizes stochastic parallel decoding for fast generation
Low Memory Footprint
Optimized memory usage, suitable for resource-limited environments
Random Order Generation
Supports image generation in random orders
Zero-shot Inference
Performs inference without additional training

Model Capabilities

Unconditional Image Generation
Class-conditional Image Generation
High-quality Image Synthesis

Use Cases

Creative Design
Art Creation
Generates unique artistic works
High-quality artistic images
Data Augmentation
Training Data Expansion
Generates training data for machine learning models
Diverse training samples
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase