--- license: apache-2.0 base_model: - Wan-AI/Wan2.2-T2V-A14B - Wan-AI/Wan2.2-I2V-A14B library_name: diffusers tags: - video_generation - NVFP4 - Sparse_Attention - Wan --- # šŸŽ¬ LightWan2.2-A14B > **An extremely efficient Wan 2.2 14B variant: NVFP4 Quantization-Aware Step Distillation with Sparse Attention for Blackwell Architecture** [![GitHub](https://img.shields.io/badge/GitHub-ModelTC/LightX2V-blue)](https://github.com/ModelTC/LightX2V) [![HuggingFace](https://img.shields.io/badge/HuggingFace-LightWan2.2--A14B-yellow)](https://huggingface.co/lightx2v/LightWan2.2-A14B) [![Blog](https://img.shields.io/badge/Blog-LightWan2.2--A14B-green)](https://light-ai.top/LightX2V-BLOG/posts/LightWan22-A14B/) ## šŸ“‹ Table of Contents - [āœØ Features](#-features) - [šŸš€ Quick Start](#-quick-start) - [šŸŽ¬ Generation Results](#-generation-results) - [āš” Performance Comparison](#-performance-comparison) - [āš ļø Notes](#ļø-notes) - [šŸ¤ Community](#-community) ## āœØ Features - **āš” 4-Step Inference**: Two high-noise expert steps followed by two low-noise expert steps, enabling extremely fast Wan2.2 MoE generation on a single Blackwell GPU. - **šŸŽÆ NVFP4 Quantization**: Quantization-aware step distillation reduces memory traffic and compute cost while targeting Blackwell architecture. - **šŸ§© Sparse Attention**: Accelerates the costly O(nĀ²) self-attention workload with sparse attention, reducing end-to-end latency for high-resolution video generation. - **šŸ”§ LightX2V Integration**: Recommended runtime stack for stable deployment and best performance. - **šŸš€ High-Quality Generation**: Preserves the visual quality of Wan2.2-T2V/I2V-14B while dramatically improving inference speed. ## šŸš€ Quick Start We strongly recommend using the official LightX2V Docker image for the cleanest environment and best reproducibility. ### Option A: Docker Recommended ```bash # 1. Pull LightX2V Docker image docker pull lightx2v/lightx2v:26052801-cu130-5090 # 2. Run single-GPU inference # Text-to-video bash scripts/wan22/extreme/run_wan22_moe_t2v_extreme.sh # Image-to-video bash scripts/wan22/extreme/run_wan22_moe_i2v_extreme.sh # 3. Run multi-GPU sequence-parallel inference # Text-to-video bash scripts/wan22/extreme/run_wan22_moe_t2v_extreme_sp_parallel.sh # Image-to-video bash scripts/wan22/extreme/run_wan22_moe_i2v_extreme_sp_parallel.sh ``` ### Option B: Manual Installation If Docker is not available, install the environment manually: ```bash # 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 single-GPU inference # Text-to-video bash scripts/wan22/extreme/run_wan22_moe_t2v_extreme.sh # Image-to-video bash scripts/wan22/extreme/run_wan22_moe_i2v_extreme.sh # 4. Run multi-GPU sequence-parallel inference # Text-to-video bash scripts/wan22/extreme/run_wan22_moe_t2v_extreme_sp_parallel.sh # Image-to-video bash scripts/wan22/extreme/run_wan22_moe_i2v_extreme_sp_parallel.sh ``` **Single-GPU Scripts:** - [run_wan22_moe_t2v_extreme.sh](https://github.com/ModelTC/LightX2V/blob/main/scripts/wan22/extreme/run_wan22_moe_t2v_extreme.sh) - [run_wan22_moe_i2v_extreme.sh](https://github.com/ModelTC/LightX2V/blob/main/scripts/wan22/extreme/run_wan22_moe_i2v_extreme.sh) **Multi-GPU Scripts:** - [run_wan22_moe_t2v_extreme_sp_parallel.sh](https://github.com/ModelTC/LightX2V/blob/main/scripts/wan22/extreme/run_wan22_moe_t2v_extreme_sp_parallel.sh) - [run_wan22_moe_i2v_extreme_sp_parallel.sh](https://github.com/ModelTC/LightX2V/blob/main/scripts/wan22/extreme/run_wan22_moe_i2v_extreme_sp_parallel.sh) ## šŸŽ¬ Generation Results

"Two anthropomorphic cats in comfy boxing gear and bright gloves fight intensely on a spotlighted stage"

| Resolution | Wan2.2-T2V-14B | LightWan2.2-A14B | | --- | --- | --- | | 480p | | | | 720p | | | ## āš” Performance Comparison **Test Environment**: RTX 5090 Single GPU | LightX2V Framework | End-to-End Latency | Method | Task | GPU Number | Resolution | NFE | E2E Latency | Speedup | | --- | :---: | ---: | ---: | ---: | ---: | ---: | | Wan2.2-T2V-14B | T2V | 1 | 480p | 40 | 734.0s | 1.0x | | LightWan2.2-A14B | T2V | 1 | 480p | 4 | 9.1s | 80.7x | | Wan2.2-T2V-14B | T2V | 1 | 720p | 40 | 2668.0s | 1.0x | | LightWan2.2-A14B | T2V | 1 | 720p | 4 | 22.5s | 118.7x | | Wan2.2-I2V-14B | I2V | 1 | 480p | 40 | 787.0s | 1.0x | | LightWan2.2-A14B | I2V | 1 | 480p | 4 | 10.7s | 73.9x | | Wan2.2-I2V-14B | I2V | 1 | 720p | 40 | 2685.0s | 1.0x | | LightWan2.2-A14B | I2V | 1 | 720p | 4 | 26.7s | 100.5x | ## āš ļø Notes ### System Requirements - **Required Hardware**: NVIDIA RTX 50-series GPUs or other Blackwell architecture GPUs. - **Recommended Runtime**: `lightx2v/lightx2v:26052801-cu130-5090`. ### Dependencies - Prepare Wan2.2 T5 / VAE components following the standard LightX2V Wan2.2 model structure. - For I2V, also prepare the required image encoder components and input image according to the LightX2V Wan2.2 I2V script. - Use Blackwell + NVFP4 kernels for optimal speed and memory efficiency. ### Performance Tips - Use the provided extreme inference script for the 4-step high-noise / low-noise expert schedule. - Sparse attention is most beneficial at higher resolutions where self-attention dominates latency. - Enable CPU offload only when GPU memory is limited, since offload can reduce throughput. ## šŸ¤ Community - **šŸ› Issues**: [GitHub Issues](https://github.com/ModelTC/LightX2V/issues) - **šŸ¤— Models**: [HuggingFace Hub](https://huggingface.co/lightx2v/) - **šŸ“– Documentation**: [LightX2V Docs](https://github.com/ModelTC/LightX2V) ---
**If you find this project helpful, please give us a ā­ on [GitHub](https://github.com/ModelTC/LightX2V)** For questions or issues, please open an issue on [LightX2V](https://github.com/ModelTC/LightX2V/issues) or contact lvchengtao0319@gmail.com.