stablediffusion
stablediffusion copied to clipboard
Stability-AI
High-Resolution Image Synthesis with Latent Diffusion Models
Hi Stability-AI team😀 I have a question. I would appreciate it if you could answer. I want to use the Image Upscaling with Stable Diffusion function. However, when I checked...
## Problem  Getting OOM memory errors at the decode stage with baseline arguments from the readme, specifying only `--ckpt "..." --prompt "..." --config "..." --H 768 --W 768` Even...
with current technology would it be possible to **ask the AI to generate a text from an image**? in order to know what technology could describe the image, a tool...
For `FrozenOpenCLIPEmbedder`, `open_clip.create_model_and_transforms()` simply calls `create_model()` with the same arguments. https://github.com/mlfoundations/open_clip/blob/9d31b2ec4df6d8228f370ff20c8267ec6ba39383/src/open_clip/factory.py#L201 We can skip the unnecessary step.
Normally the CUDA oom is a normal thing with smaller GPUs but... 400GiB? I dont think that exists as a gpu so this is obviously a bug. 512x512 input. Goes...
Hi, thanks for you great job ! when i run txt2img.py, it broken and killed. Do you know what's the problem ? DiffusionWrapper has 865.91 M params. making attention of...
Using the 768 v-diffusion model, using prompt "fruit basket". With DDIM sampling:  With PLMS sampling: 
## Background Hello there! The Python packaging ecosystem has standardized on the interface for build backends ([PEP 517](https://peps.python.org/pep-0517/)/[PEP 660](https://peps.python.org/pep-0660/)) and the format for metadata declaration ([PEP 621](https://peps.python.org/pep-0621/)/[PEP 631](https://peps.python.org/pep-0631/)). As a...
Automatic code formatting reduces mental overhead for developers. - Code style is consistent in this repo and across other repos - Future diffs will not have anything relating to style...
When you run `conda env create`, conda automatically looks for `environment.yml`. With `.yaml`, you will have to specify the file name manually.
Metadata
Owner
Metadata
High-Resolution Image Synthesis with Latent Diffusion Models