Automatic1111 Instance Setup
Here are some setups that I use to load the models I use after setting up an instance. I would ssh
into the instance and run them. Inside the instance, the path /workspace
is where you can read/write. It’s also where the stable diffusion loads models and outputs your images. You have root access, so you can install anything via apt install
. sudo
is not required since you’re already logged in as root
.
Civitai Downlaods
Some of these download from Civitai directly. A few notes:
- Download URL.
- You can’t download the URL to the model.
- You must find the download link on the model page, which has
/api
in the path. - For most models that don‘t have multiple versions, it’s the same link as the button.
- Right-click on that button to copy the
/api
link or I have done it already for you below.
--content-disposition
- This is required to download the file with the correct name.
- Otherwise, it will be downloaded as some filename that matches the URL instead, and you will have to rename it.
-O
- If you want to save the file in a different name, you can use the
-O
(that’s uppercaseO
) followed by the filename. - You rarely have to do this.
- If in doubt, use the
--content-disposiion
option. - Some models on Civitai require it but it’s like 1% of them.
- If you want to save the file in a different name, you can use the
Civitai API
Civitai provides a comprehensive API. For nerds out there who want to automate a lot of these without copying and pasting, you can use their API to write almost everything on this page with code. Here are their docs:
Civitai Discord
For convenience, here’s a link to their Disccord server. They are fairly responsive in their dev channel:
Models
SDXL
# SDXL 1.0 Base
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/resolve/main/sd_xl_base_1.0.safetensors
# SDXL 1.0 Refiner
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0/resolve/main/sd_xl_refiner_1.0.safetensors
# Fixed 0.9 VAE
cd /workspace/stable-diffusion-webui/models/VAE
wget https://huggingface.co/madebyollin/sdxl-vae-fp16-fix/resolve/main/sdxl_vae.safetensors
# SDXL Offset Noise LoRA
cd /workspace/stable-diffusion-webui/models/Lora
wget https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/resolve/main/sd_xl_offset_example-lora_1.0.safetensors
# 4x_NMKD-Siax_200k upscaler
cd /workspace/stable-diffusion-webui/models/ESRGAN
wget https://huggingface.co/uwg/upscaler/resolve/main/ESRGAN/4x_NMKD-Siax_200k.pth
# 4x-UltraSharp upscaler
cd /workspace/stable-diffusion-webui/models/ESRGAN
wget https://huggingface.co/uwg/upscaler/resolve/main/ESRGAN/4x-UltraSharp.pth
A-Zovya Photoreal 2
General purpose photorealistic SFW checkpoint model
https://civitai.com/models/57319?modelVersionId=99805
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/99805 --content-disposition
Airfuck’s Wild Mix
Male-focus NSFW checkpoint model
https://civitai.com/models/82075?modelVersionId=87131
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/87131 --content-disposition
Airfuck’s Brute Mix
https://civitai.com/models/135202/airfucks-brute-mix
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/148997 --content-disposition
CyberRealistic
Photorealistic checkpoint model - general purpose SFW.
https://civitai.com/models/15003/cyberrealistic
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/109054 --content-disposition
3.3
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget 'https://civitai.com/api/download/models/138176?type=Model&format=SafeTensor&size=pruned&fp=fp16' --content-disposition
# CyberRealistic Negative Embedding
cd /workspace/stable-diffusion-webui/embeddings
wget 'https://civitai.com/api/download/models/82745?type=Negative&format=Other' --content-disposition
Juggernaut
Photorealistic checkpoint model.
Aftermath
https://civitai.com/models/46422?modelVersionId=127207
Original version. juggernaut_aftermath.safetensors
# 5.28 GB
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget 'https://civitai.com/api/download/models/127207?type=Model&format=SafeTensor&size=full&fp=fp32' -O juggernaut_aftermath.safetensors
Pruned version. They are the same. Pruned is simply a smaller file that doesn’t have all the unused data. juggernaut_aftermath.safetensors
# 1.99 GB - Pruned
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget 'https://civitai.com/api/download/models/127207?type=Model&format=SafeTensor&size=pruned&fp=fp16' -O juggernaut_aftermath.safetensors
Config needs to be saved to the same directory as the model. juggernaut_aftermath.yaml
# 1.99 GB - Pruned
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget 'https://civitai.com/api/download/models/127207?type=Config&format=Other' -O juggernaut_aftermath.yaml
VAE needs to be in the VAE folder TRCVAE.safetensors
cd /workspace/stable-diffusion-webui/models/VAE
wget 'https://civitai.com/api/download/models/127207?type=VAE&format=SafeTensor' -O TRCVAE.safetensors
Juggernaut XL
https://civitai.com/models/133005
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget 'https://civitai.com/api/download/models/146359' --content-disposition
Virile Animation 1
Male-focus NSFW checkpoint model for a 2d anime look
https://civitai.com/models/101879?modelVersionId=109054
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/109054 --content-disposition
Virile Fantasy
https://civitai.com/models/111238?modelVersionId=120374
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/120374 --content-disposition
Virile Fusion 2
Male-focus NSFW checkpoint model for a photorealistic 3d look
https://civitai.com/models/77043?modelVersionId=98297
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/98297 --content-disposition
Virile Reality 2
Male-focus NSFW checkpoint model for photorealistic photogarphy
https://civitai.com/models/82790?modelVersionId=105157
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/105157 --content-disposition
Virile Motion 1
Male-focus NSFW checkpoint model for a pixar animation look
https://civitai.com/models/90813?modelVersionId=96763
cd /workspace/stable-diffusion-webui/models/Stable-diffusion
wget https://civitai.com/api/download/models/96763 --content-disposition
VAE
Remember to change the VAE setting in the WebUI Settings!
vae-ft-ema-560000-ema-pruned.ckpt
cd /workspace/stable-diffusion-webui/models/VAE
wget https://huggingface.co/stabilityai/sd-vae-ft-ema-original/resolve/main/vae-ft-ema-560000-ema-pruned.ckpt -O vae-ft-ema-560000-ema-pruned.ckpt
vae-ft-ema-840000-ema-pruned.ckpt
cd /workspace/stable-diffusion-webui/models/VAE
wget https://huggingface.co/stabilityai/sd-vae-ft-mse-original/resolve/main/vae-ft-mse-840000-ema-pruned.ckpt -O vae-ft-mse-840000-ema-pruned.ckpt
kl-f8-anime2-vae
VAE for anime
https://civitai.com/models/23906/kl-f8-anime2-vae
cd /workspace/stable-diffusion-webui/models/VAE
wget 'https://civitai.com/api/download/models/28569?type=VAE&format=Other' -O MoistMix.vae.pt
Control Net
WebUI docker image on vast only comes with Canny, so we’ll have to download the models from Hugging Face. There are two ways: either clone the entire model repo (40GB) or download indivdually. Unless you need all the models, it’s faster and require less disk space to just get the models you need.
Clone entire repo
Install git LFS
https://github.com/git-lfs/git-lfs/blob/main/INSTALLING.md
mkdir /workspace/tmp
cd /workspace/tmp
curl -s https://packagecloud.io/install/repositories/github/git-lfs/script.deb.sh | sudo bash
apt-get install git-lfs
Clone the Repo
This should work but it’s super slow. Also, the full model git repo is currently at 40 GB. So instead of this, just download via Hugging Face API below
mkdir /workspace/downloads
cd /workspace/downloads
git lfs install
git clone https://huggingface.co/lllyasviel/ControlNet-v1-1
Download via wget
- Quality flags: https://github.com/lllyasviel/ControlNet-v1-1-nightly
- ControlNet 1.1 include 14 models (11 production-ready models and 3 experimental models):
CN Tile
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11f1e_sd15_tile.pth
CN Soft Edge
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11p_sd15_softedge.pth
CN Open Pose
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11p_sd15_openpose.pth
CN Line art
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11p_sd15_lineart.pth
CN Shuffle
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11e_sd15_shuffle.pth
CN Depth
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11f1p_sd15_depth.pth
CN Inpaint
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11p_sd15_inpaint.pth
CN line art anime
cd /workspace/stable-diffusion-webui/extensions/sd-webui-controlnet/models
wget https://huggingface.co/lllyasviel/ControlNet-v1-1/resolve/main/control_v11p_sd15s2_lineart_anime.pth
Textual Inversions
# Textual Inversions
cd /workspace/stable-diffusion-webui/embeddings
# NG Deep Negative
wget https://civitai.com/api/download/models/5637 --content-disposition
# EasyNegative
wget https://civitai.com/api/download/models/9208 --content-disposition
# NegFeetv2
wget https://civitai.com/api/download/models/98441 --content-disposition
# Asian-Less
wget https://civitai.com/api/download/models/57451 --content-disposition
# Bad Hands
wget https://civitai.com/api/download/models/125849 --content-disposition
# GS Masculine
wget https://civitai.com/api/download/models/95532 --content-disposition
LoRA
Add Detail
Adding detail for txt2img
https://civitai.com/models/58390/detail-tweaker-lora-lora
cd /workspace/stable-diffusion-webui/models/Lora
wget 'https://civitai.com/api/download/models/62833?type=Model&format=SafeTensor' --content-disposition
Gender Slider
Some LoRA are trained on a different gender, e.g. female, and it will turn your renders into female. This LoRA will correct that. For gay gens, use the Masculine version.
https://civitai.com/models/89709?modelVersionId=95532
cd /workspace/stable-diffusion-webui/models/Lora
wget 'https://civitai.com/api/download/models/95532' --content-disposition
Zoom Slider
Are you sick up having to write full body
, close up
, medium shot
, wide angle
and somehow they never really always work? The Zoom Slider LoRA will let you frame your subject from the correct distance perfectly. I use it all the time. It really works!
https://civitai.com/models/114460?modelVersionId=123732
cd /workspace/stable-diffusion-webui/models/Lora
wget 'https://civitai.com/api/download/models/123732?type=Model&format=SafeTensor' -O zoom_slider_v1.safetensors
Bulgerk-dickprint
Prevent bulge from being just a “balloon” and have some shape to them. I use this sometimes but honestly most of the gay models already do it perfectly. These are trained with real bulges. Made by the author who made the bulgerk
series.
https://civitai.com/models/102191?modelVersionId=109392
cd /workspace/stable-diffusion-webui/models/Lora
wget 'https://civitai.com/api/download/models/109392' --content-disposition
epi_noiseoffset
For controlling lighting in the scene. Extremely helpful to create dark scenes with minimal light.
https://civitai.com/models/13941/epinoiseoffset
cd /workspace/stable-diffusion-webui/models/Lora
wget 'https://civitai.com/api/download/models/16576?type=Model&format=SafeTensor&size=full&fp=fp16' --content-disposition
Hypermuscle
Don’t usually need this if you’re working with gay models, but useful if you want to add muscles to a straight model.
https://civitai.com/models/8084/hypermuscle
cd /workspace/stable-diffusion-webui/models/Lora
wget 'https://civitai.com/api/download/models/9537' --content-disposition
Male feet pose
https://civitai.com/models/53688/male-feet-pose
cd /workspace/stable-diffusion-webui/models/Lora
wget 'https://civitai.com/api/download/models/83769' --content-disposition
WebUI extensions
You could use the GUI to install these extensions, but it‘s much faster to:
cd
into the extension foldergit clone
all the extensions you need- Restart the WebUI process with
fuser -k 3000/tcp
- Wait a few seconds and you should be good to go
Installing extensions this way is what the WebUI does behind the scene. (Source: Automatic1111, Runpod)
fuser -k 3000/tcp
ADetailer
https://github.com/Bing-su/adetailer
https://github.com/Bing-su/adetailer
cd /workspace/stable-diffusion-webui/extensions
git clone https://github.com/Bing-su/adetailer.git
Face Editor
(In place of ADetailer)
https://github.com/ototadana/sd-face-editor
cd /workspace/stable-diffusion-webui/extensions
git clone https://github.com/ototadana/sd-face-editor.git
Dynamic Prompts
https://github.com/adieyal/sd-dynamic-prompts
https://github.com/adieyal/sd-dynamic-prompts
cd /workspace/stable-diffusion-webui/extensions
git clone https://github.com/adieyal/sd-dynamic-prompts.git
Regional Prompter
https://github.com/hako-mikan/sd-webui-regional-prompter
https://github.com/hako-mikan/sd-webui-regional-prompter
cd /workspace/stable-diffusion-webui/extensions
git clone https://github.com/hako-mikan/sd-webui-regional-prompter.git
ROOP
https://github.com/s0md3v/sd-webui-roop
https://github.com/s0md3v/sd-webui-roop
ROOP NSFW
https://github.com/Gourieff/sd-webui-roop-nsfw
https://github.com/Gourieff/sd-webui-roop-nsfw
SD Ultimate Upscale
https://github.com/Coyote-A/ultimate-upscale-for-automatic1111
https://github.com/Coyote-A/ultimate-upscale-for-automatic1111
Upscalers
https://upscale.wiki/wiki/Model_Database
https://upscale.wiki/wiki/Official_Research_Models
4x_foolhardy_Remacri
cd /workspace/stable-diffusion-webui/models/ESRGAN
wget https://huggingface.co/FacehugmanIII/4x_foolhardy_Remacri/resolve/main/4x_foolhardy_Remacri.pth
4x_NMKD-Siax_200k
cd /workspace/stable-diffusion-webui/models/ESRGAN
wget https://huggingface.co/Akumetsu971/SD_Anime_Futuristic_Armor/resolve/main/4x_NMKD-Siax_200k.pth