interactdiffusion

interactdiffusion copied to clipboard

OUT OF MEMORY

1

Why does my training run out of memory when the batch size is set to 4, and out of memory when the batch size is set to 2 during multi-GPU...

BlackPuuuudding

Thank you for your great work! Will you release the weights of GLIGEN*? I believe this will greatly foster future research on this topic.

daoyuan98

Thanks for the impressive work. I try the webui implementation but I find the generated image cannot follow the defined bounding boxes. How should I fix this issue. Prompt: a...

zzhang2816

SDXL Compatibility

1

The concept suggested in this paper and the corresponding trained model work very well with regards to the scope of the paper, which is enabling object/subject interaction generation. However, due...

rohanath123

Custom Dataset

3

Thank you for sharing your research in this code. Thanks to you, I am studying a lot. I want to conduct an experiment on custom dataset. If you look at...

Hayeon-kimm

The loss doesn't decrease

3

Hello, I'm trying to train InteractDiffusion using the HOI dataset according to the readme, but I've noticed that the loss isn't decreasing. Instead, it's fluctuating. Is this normal? ![截屏2024-07-03 16...

hugo-xie

Parameter has no gradient

1

BlackPuuuudding

Inference Error

4

When I use the command `python inference_batch.py --batch_size 1 --folder generated_output --seed 489 --scheduled-sampling 1.0 --half`, unexpected error occurs: ``` RuntimeError: Error(s) in loading state_dict for FrozenCLIPEmbedder: Unexpected key(s) in...

Hammour-steak

fine-tuning GLIGEN*

2

For fine-tuning GLIGEN*, around which step does it typically converge? This information would be helpful for my further research work. Are the parameter settings consistent?

Mian2001