daam

daam copied to clipboard

can use the method to show the cross attention map in the stable-cascade?

Joyies

Support for Prompt Embeddings Input Argument during Inference

4

This is more of a request, but would you be able to support using custom embeddings and negative embeddings as pipeline arguments? The reason I want to do this is...

chrisprasanna

Hi, Thanks for this great paper+codebase! A deserved best paper award. I am thinking of applying the DAAM idea to image editing models (i.e. InstructPix2Pix: https://github.com/timothybrooks/instruct-pix2pix) and was wondering how...

BennoKrojer

Showing <|startoftext|> and <|endoftext|>

1

Hi, Thanks for a great work in visualising the cross attention of diffusion models. Is it possible to visualise the cross-attention map for the tokens of and ? Since they...

jiuntian

Support for self attention visualization

1

Hi, I find your work very interesting. Is there a way to visualize self attention maps also in the existing repo?

thakral-kartik

How to try semantic segmentation on real images

5

I want to try segmentation on several images, can you give me some assistance.

HanLiii

This is not an issue but a request. Can we run this on any image with generated captions?

3

This will help us in creating better descriptions of an image while training Textual inversion (embeddings) and LoRA. How to use this on any image? **does it work other-way around...

runner22k

When I use it with `stabilityai/stable-diffusion-2-inpainting` I get this error below. 

himmetozcan

Add support for non-square output

1

Hi, Thanks for the great work! I wonder if you can add the support for non-square image outputs? For example, the output height is 704 and width is 512. That...

ArmastusChen

Support for Stable Diffusion 2 (non-base) | 768x768 tensor size

6

Title. DAAM currently (as per 0.9.0) only supports SD2 512x512 while the original model has 768x768 sizes.

federicotorrielli