Downloads · 30 days
0
wan22-loras
WindstormLabs/wan22-loras
wan22-loras is a machine learning model from WindstormLabs. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Organization: WindstormLabs Used by: SceneMachine
Access
Public
Updated Oct 1, 2026
Repo size
1.9 GB
Likes
0
Public
Repo makeup
Click a slice to open those files.
.safetensors1.9 GB · 100%
At a glance
- License
- apache-2.0
- Access
- Public
- Created
- May 14, 2026
- Updated
- Oct 1, 2026
- SHA
- 0bf5ccdb
Model card
From the Hugging Face model README
Wan 2.2 Lightx2v 4-step distillation LoRAs
Organization: WindstormLabs
Used by: SceneMachine
SceneMachine is a sub-project of Windstorm Labs. This repo hosts shared AI/ML infrastructure used by SceneMachine and reusable by future Windstorm Labs sub-projects.
What this is
Kijai's Lightx2v 4-step distillation LoRAs for Wan 2.2. Drops sampler from 30 steps to 4 with cfg=1.0 — 8.3× wallclock speedup measured on Wan Animate (844s → 102s/shot). Validated to transfer cleanly to Animate when paired with the correct embed chain + CLIP-ViT-H.
Upstream source
Primary distribution: https://huggingface.co/Kijai/WanVideo_comfy/tree/main/LoRAs/Wan22_Lightx2v + https://huggingface.co/Comfy-Org/Wan_2.2_ComfyUI_Repackaged
This repo is a mirror. License terms of the upstream apply unchanged.
Primary license owner(s): Kijai (LoRA) / Alibaba (base model Apache-2.0)
Files
| Filename | Size |
|---|---|
wan2.2_i2v_lightx2v_4steps_lora_v1_high_noise.safetensors | 1.23 GB |
Wan_2_2_I2V_A14B_HIGH_lightx2v_4step_lora_260412_rank_64_fp16.safetensors | 0.63 GB |
Total: 1.86 GB
Related repos
- The full SceneMachine model stack: search the SceneMachine HF collection.
- Shared encoders and Wan VAE:
WindstormLabs/wan22-encoders. - Speed LoRAs:
WindstormLabs/wan22-loras.
Licence
Apache-2.0, inherited from the upstream Wan 2.2 release by Wan-AI (Alibaba). Licence text: LICENSE.
Provenance
Mirror created 2026-05-13 from a local working copy. Hashes preserved via HF's content-addressed storage.
- License
- apache-2.0
- Languages
- en
- Created
- May 14, 2026
- Updated
- Oct 1, 2026