Downloads · 30 days
6
2% of all-time downloads
Wan2.2-S2V-14B-overlay
MickJ/Wan2.2-S2V-14B-overlay
Wan2.2-S2V-14B-overlay is a machine learning model from MickJ. Use it for the machine learning task on the model card, and read the license before you ship it in a product. It is set up for diffusers.
This is a metadata-only overlay repo for Wan-AI/Wan2.2-S2V-14B.
All-time downloads
378
Public
Repo size
—
Likes
1
Public
Repo makeup
Click a slice to open those files.
.py2.5 KB · 46%
At a glance
- Library
- diffusers
- Access
- Public
- Created
- Mar 28, 2026
- Updated
- Mar 29, 2026
- SHA
- 355f9aed
Model card
From the Hugging Face model README
Wan2.2-S2V-14B Overlay
This is a metadata-only overlay repo for Wan-AI/Wan2.2-S2V-14B.
It contains SGLang pipeline metadata plus vendored official wan/ code. Model
weights are still downloaded from the original source repo at runtime and then
materialized into a local componentized cache.
Current component classes:
transformer:WanS2VTransformer3DModelscheduler:FlowUniPCMultistepSchedulertext_encoder:UMT5EncoderModelaudio_encoder:WanS2VAudioEncodervae:AutoencoderKLWan
The DIT entry in model_index.json is native SGLang (WanS2VTransformer3DModel),
not the old WanModel_S2V / WanS2VOfficial* naming.
This overlay also bumps materializer_version so an existing local
materialized cache will be rebuilt with the native component names.
Use directly:
sglang generate --model-path MickJ/Wan2.2-S2V-14B-overlay --config configs/wan_s2v.yaml
Or map the official source model ID through SGLANG_DIFFUSION_MODEL_OVERLAY_REGISTRY.
- Library
- diffusers
- Created
- Mar 28, 2026
- Updated
- Mar 29, 2026