OpenImagingLab / OpenImagingLab/FlashVSR
about train stage 1
Open
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 1.9k
- Forks
- 152
- PR merge metrics
- No merged PRs in 30d
Description
作者您好，论文中提到训练是分3阶段完成的，在第三阶段因为变成了单步dit，所以可以引入感知loss，那么请问在第一阶段训练时，由于只有flow matching loss，该阶段的网络对于视频细节还原能足够好吗？
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
The issue names no files, tests, or entry points. Start by reviewing the repository's training configuration and any documentation describing the three training stages, then compare the stated losses with the paper; completion would require an authoritative explanation of whether stage one preserves sufficient video detail.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- computer-vision, machine-learning
- Issue type
- Documentation
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 15/100