aigc-apps / aigc-apps/VideoX-Fun
关于CogVideoX-Fun-V1.1-2b-InP模型
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 2.3k
- Forks
- 188
- Avg merge
- 2d 22h
- Merged PRs (30d)
- 3
Description
这个模型是基于cogvideox的i2v模型微调得到的吗，但我看其中tranformers的部分的in_channels是33，但原始模型是32，具体是怎么基于原始模型微调训练的呢，还是完全从头训练了。
还有个问题，在i2v中，在对输入的视频进行vae编码之前，是把非首帧的值设置为了-1，但感觉设置成0更好一些？
masked_video = init_video * (mask_condition_tile < 0.5) + torch.ones_like(init_video) * (mask_condition_tile > 0.5) * -1
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 comparing the CogVideoX-Fun-V1.1-2b-InP transformer configuration with the original CogVideoX configuration, then trace the I2V training entry point to determine how the in_channels difference is handled. Inspect the masked_video preprocessing expression and document whether the model was fine-tuned or trained from scratch and why non-first frames use -1 rather than 0.
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
- 25/100