Stable Diffusion XL
POST ON 2024-02-01 BY WOLVES
About
Stable Diffusion is a latent text-to-image diffusion model. Thanks to a generous compute donation from Stability AI and support from LAION, we were able to train a Latent Diffusion Model on 512x512 images from a subset of the LAION-5B database. Similar to Google’s Imagen, this model uses a frozen CLIP ViT-L/14 text encoder to condition the model on text prompts. With its 860M UNet and 123M text encoder, the model is relatively lightweight and runs on a GPU with at least 10GB VRAM. See this section below and the model card.
Stable Diffusion Web UI
项目地址
github.com/AUTOMATIC1111/stable-diffusion-webui
常用模型/提示词网站
Stable Diffusion Web UI On Linux
1.Launch webui.sh
- For ubuntu, Please use user instead of root
- webui will install requirements by itself in directory venv , which includes with downloading basic model2.If you find the red line: Cannot locate TCMalloc. Do you have tcmalloc or google-perftool installed on your system? (improves CPU memory usage)
- For ubuntu, Please test `apt-get install libgoogle-perftools4 libtcmalloc-minimal4 -y`3.Because in linux without public Ipv4, you may use the proxy to access webui
- Please write args in webui-user.sh
- --listen --enable-insecure-extension-access --no-gradio-queue --share 4.There are some args else
- To set user and password for protect, you can use `--gradio-auth user:passwd`
- To promote perfermence, you can use `--xformers` for enable xformers` support
- To select GPU, you can use `--device-id 0` to set, the number is gpu index, and you should know sd is a text-to-image by one gpu, not design for muti-gpusERROR List