🎬 Self-Forcing-NVFP4-4Steps Models

NVFP4 Quantization-Aware Step Distillation for Blackwell Architecture

📋 Table of Contents

• ✨ Features
• 🚀 Quick Start
• 🎬 Generation Results
• 📦 Installation
• 🛠️ Usage
• 🧭 Project Structure
• ⚠️ Notes
• 🤝 Community

✨ Features

• ⚡ 4-Step Inference: Dramatically accelerated end-to-end generation approaching real-time performance (tested on RTX 5090 single GPU)
• 🎯 NVFP4 Quantization: Reduced memory and bandwidth usage, optimized for Blackwell architecture
• 🔧 LightX2V Integration: Optimal performance and stability on the official framework
• 🚀 High-Quality Generation: Maintains Self-Forcing's superior video quality while achieving unprecedented speed

🚀 Quick Start

# 1. Install LightX2V
git clone https://github.com/ModelTC/LightX2V.git
cd LightX2V
uv pip install -v .

# 2. Install NVFP4 Kernel
pip install scikit_build_core uv
git clone https://github.com/NVIDIA/cutlass.git
cd lightx2v_kernel

MAX_JOBS=$(nproc) CMAKE_BUILD_PARALLEL_LEVEL=$(nproc) \
uv build --wheel \
  -Cbuild-dir=build . \
  -Ccmake.define.CUTLASS_PATH=/path/to/cutlass \
  --verbose --color=always --no-build-isolation

pip install dist/*whl --force-reinstall --no-deps

# 3. Run inference
# config
https://github.com/ModelTC/LightX2V/blob/main/configs/self_forcing/wan_t2v_sf_nvfp4.json

🎬 Generation Results

"A leprechaun, with green hat and traditional Irish attire, standing in a lush forest filled with vib..."

"A mystical and spiritual scene filled with loving energy emanating from the heavens. The sky is bath..."

⚠️ Notes

System Requirements

• Required Hardware: NVIDIA RTX 50-series GPUs (RTX 5090/5080/5070/5060) or other Blackwell architecture GPUs

Dependencies

• Prepare T5 / CLIP / VAE components yourself (same as Self-Forcing structure)

Performance Tips

• Use Blackwell + NVFP4 for best performance
• Enable CPU offload for GPUs with limited memory

🤝 Community

• 🐛 Issues: GitHub Issues
• 🤗 Models: HuggingFace Hub
• 📖 Documentation: LightX2V Docs

---

If you find this project helpful, please give us a ⭐ on GitHub