-
Hi there,
Neat work!
I had a question if this codebase would support generating a novel view using a single image?
-
### Project Title: Image and Audio-Driven Video Generation
**Description:**
This project focuses on creating a system that generates a realistic video from a single image and audio input. By fee…
-
How Do I Make it Generate single images instead of multiple ones (usually 4)
-
Despite the DDPM limitations you mentioned, there still is the possibility to do multi-image conditioning. Has any work been done concerning multi-image2singleimage generation with the BBDM framework?
-
Hi,
When i try to run text-to-image generation with single A100 40GB, it is out of memory at the decoding phase (with VQModel). Is that unusual ?
Thanks
-
-
Thanks for your awesome work! May I ask when the video-version world dreamer would be released? Thanks!
-
**Describe the bug**
When using the PVP Style feature (for single target, aoe or for winder combo), the reawaken feature only displays the first generation/legacy combo, and gets stuck at that. It ne…
-
# Single-Device-Abstract DDP
## Motivation
In current PyTorch DDP, when training a model with Dropout operations, the final results obtained from distributed training will not be consistent with t…
-
### Motivation
I notice that internval_chat/eval/evaluate_vqa.py has parameters for few-shot learning but have not been implemented correctly.
My question is:
How can we do few-shot learning …