GRN: Generative Refinement Networks

This is the official implementation of the paper Generative Refinement Networks for Visual Synthesis. Neither diffusion nor autoregressive — GRN is a third way. 🧠 Refines globally like an artist. ⚡ Generates adaptively by complexity. 🏆 New SOTA across image & video. The visual generation paradigm just got rewritten.

---

📋 Table of Contents

• 🌟 Introduction
• 📑 Open-Source Plan
• 🛠️ Installation
• 🖼️ Class-to-Image
  • Dataset
  • Training
  • Evaluation
• 📧 Contact
• 🤗 Acknowledgements
• 📝 Citation

---

🌟 Introduction

Diffusion models dominate visual generation but they allocate uniform computational effort to samples with varying levels of complexity. Autoregressive (AR) models are complexity-aware, as evidenced by their variable likelihoods, but suffer from lossy tokenization and error accumulation.

We introduce Generative Refinement Networks (GRN), a new visual synthesis paradigm that addresses these issues:

• Near-lossless tokenization via Hierarchical Binary Quantization (HBQ)
• Global refinement mechanism that progressively perfects outputs like a human artist
• Entropy-guided sampling for complexity-aware, adaptive-step generation

GRN achieves state-of-the-art results on ImageNet reconstruction and class-conditional generation, and scales effectively to text-to-image and text-to-video tasks.

Generative Refinement Framework

Starting from a random token map, GRN randomly selects more predictions at each step and refines all input tokens. For example, compared to the second step, the third step filled six new tokens (pink), kept two tokens (blue), erased two tokens (yellow), and left six tokens blank (gray).

Class-to-Image Examples

Text-to-Image Examples

Text-to-Video Examples

t2v_examples.mp4

---

📑 Open-Source Plan

GRN adopts a minimalist and self-contained design. This implementation is in PyTorch + GPU.

Task	Checkpoints	Inference Code	Training Code
T2V	⬜	⬜	✅
T2I	⬜	⬜	✅
C2I	⬜	✅	✅

---

🛠️ Installation

Step 1: Clone the repository

git clone https://github.com/MGenAI/GRN
cd GRN

Step 2: Create conda environment

A suitable conda environment named GRN can be created and activated with:

conda env create -f environment.yaml
conda activate GRN

Troubleshooting

If you get undefined symbol: iJIT_NotifyEvent when importing torch, simply:

pip uninstall torch
pip install torch==2.5.1 --index-url https://download.pytorch.org/whl/cu124

Check this issue for more details.

---

🖼️ Class-to-Image

Dataset

Download ImageNet dataset, and place it in your IMAGENET_PATH.

Training

All training scripts are located in scripts/c2i/. We suggest using 8x80GB GPUs for most models.

Model	Training Script	GPUs Required
GRN_ind_B	bash scripts/c2i/train_GRN_ind_B.sh	8x80GB
GRN_bit_B	bash scripts/c2i/train_GRN_bit_B.sh	8x80GB
GRN_ind_L	bash scripts/c2i/train_GRN_ind_L.sh	8x80GB
GRN_ind_H	bash scripts/c2i/train_GRN_ind_H.sh	16x80GB
GRN_ind_G	bash scripts/c2i/train_GRN_ind_G.sh	32x80GB

Evaluation

PyTorch pre-trained models are available here.

All evaluation scripts are located in scripts/c2i/. We suggest using 8x80GB vRAM GPUs.

Model	Evaluation Script
GRN_ind_B	bash scripts/c2i/eval_GRN_ind_B.sh
GRN_bit_B	bash scripts/c2i/eval_GRN_bit_B.sh
GRN_ind_L	bash scripts/c2i/eval_GRN_ind_L.sh
GRN_ind_H	bash scripts/c2i/eval_GRN_ind_H.sh
GRN_ind_G	bash scripts/c2i/eval_GRN_ind_G.sh

We use torch-fidelity to evaluate FID and IS against a reference image folder or statistics. We use the JiT's pre-computed reference stats under grn/utils_c2i/fid_stats.

---

📧 Contact

If you are interested in scaling GRN for image generation / image editing / video generation / video editing / unified model directions, please feel free to reach out!

📧 Email: hanjian.thu123@bytedance.com

---

🤗 Acknowledgements

• Thanks to JiT, Infinity and InfinityStar for their wonderful work and codebase!

---

📝 Citation

If you find our work useful, please consider citing:

@misc{han2026grn,
      title={Generative Refinement Networks for Visual Synthesis}, 
      author={Jian Han and Jinlai Liu and Jiahuan Wang and Bingyue Peng and Zehuan Yuan},
      year={2026},
      eprint={2604.13030},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2604.13030}, 
}