mihirp1998/VADER
Video Diffusion Alignment via Reward Gradients. We improve a variety of video diffusion models such as VideoCrafter, OpenSora, ModelScope and StableVideoDiffusion by finetuning them using various reward models such as HPS, PickScore, VideoMAE, VJEPA, YOLO, Aesthetics etc.
Python
Issues
- 3
How many prompts did you use in training?
#13 opened by LAW1223 - 0
Release of Baseline code
#14 opened by Sun-Haoyuan23 - 4
when to release finetinued weights
#12 opened by LAW1223 - 1
- 0
- 4
Question about the training cost.
#7 opened by yiboz2001 - 1
- 3
CUDNN_STATUS_NOT_SUPPORTED
#8 opened by yyxisgreat - 1
please add google colab or HF demo
#1 opened by s9anus98a - 2
- 2
- 2
does this work for any I2V model?
#6 opened by SkylerZheng - 1
- 0
when to release open sora finetinued weights
#2 opened by moclimb