pusa model
#46
by
rahul7star
- opened
the one uploaded here dosent work when using via code perhaps i am missing some sampling config?
What code is that? It's a LoRA for Wan14B T2V model and it requires different handling of timesteps than standard Wan inference does.
Your other models works well but when i load this pusa i get blurry video
https://huggingface.co/spaces/rahul7star/InstaVideo/blob/main/app.py (line 24)
I do get the video generated however its blurry
does it need a twek in pipeling hmm