modelscope / modelscope/DiffSynth-Studio

About Wan2.2 Vace

Open
#1,430 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
13.1k
Forks
1.3k
Avg merge
13h 12m
Merged PRs (30d)
45

Description

In the examples/wanvideo/model_inference/Wan2.2-VACE-Fun-A14B.py file, the VAE used is the Wan2.1 VAE instead of the Wan2.2 VAE. Is this normal? Why doesn't it use the Wan2.2 VAE like in examples/wanvideo/model_inference/Wan2.2-TI2V-5B.py?
Perhaps the VAE performance of Wan2.2 and 2.1 is similar, and the improvements mainly come from the DiT ?
Thank you!

Image

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Compare the VAE references in examples/wanvideo/model_inference/Wan2.2-VACE-Fun-A14B.py and examples/wanvideo/model_inference/Wan2.2-TI2V-5B.py, then trace the model-loading entry points used by each example. Confirm whether the Wan2.1 VAE choice is intentional; done means the rationale is documented or the example is aligned with the intended VAE.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Quiet
Clarity
Mostly clear
Newbie friendliness
48/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.