modelscope / modelscope/DiffSynth-Studio
wan2.1 是否支持基于多个 reference 帧的训练
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 13.1k
- Forks
- 1.3k
- Avg merge
- 13h 12m
- Merged PRs (30d)
- 45
Description
目前的 wan2.1（包括其他版本的 wan）只基于一张 reference_image 进行训练，是否支持多帧的 reference 进行训练？这部分的代码实现似乎比较简单，但看起来没有实现这个功能，是因为多帧训练表现不好吗
Contributor guide
No contributing guide indexed for this repository
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
Start by locating the wan2.1 training entry point and the code path that consumes the single reference_image. Compare this behavior with other wan versions and inspect existing training tests or examples, if present. Done would require a clear decision on multi-frame reference support, its expected behavior, and evidence about training quality or limitations.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- machine-learning
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Needs clarification
- Newbie friendliness
- 35/100