Dain Kim

First of all, thanks for the amazing project! My teammates and I are using your work in part of our course project, and we are using the FashionIQ dataset, which...

Applying Blockwise Weights to LoRA

2

It would be wonderful if I could apply stepwise or blockwise weights to LoRA, just like the webui extension below. https://github.com/hako-mikan/sd-webui-lora-block-weight Does anyone know if it's also possible in diffusers??

visualizing self attention map

3

Hi, I wonder how these figures were obtained. SVD on self-attention map would produce U, S, and V.T. How did you obtain the figures?? 

Hi, I'm still a bit confused about how you configured the valid set. To construct a valid set from the data you provide, should it be created in the test...

Hi, I tried to run the inference using the example you provided ('test_data_small'), but it returns a black empty image.

There's a bug when running local gradio demo.

requires_grad in AttnProcessor

3

Hi, thanks for the wonderful work! While looking into the code, I found that DivideBindAttnProcessor stores the attention map only if requires_grad is True. And it differs from the implementation...

I wonder if there are any guides for the hyperparameter settings. I tried object pasting on jack daniel, but the object is distorted. Are there any suggestions to preserve the...

[dataset] Collecting rankings from different annotators

1

Hello, I wonder how you collected the rankings from different annotators, especially when there's a tie. For instance, say there are six annotators, and three ranked an image as rank2...

Hello, I wonder how win/lose images are generated when constructing a PickaPic dataset. It seems like two images were generated from the same prompt, but are there any differences between...