🧩 DiffuseCraft Mod

This space is a modification of r3gm's DiffuseCraft.
Quick settings
Task
Model

You can enter a huggingface model repo_id to want to use.
Model Type
Genre
Speed
Aspect Ratio
Quality Tags Presets
Style Preset
Sampler Quick Settings
Optimization for SDXL
0 1
0 1
Convert tags to

danbooru for Animagine, e621 for Pony.
Insert reccomended prompt
Remove tags leaving only the following
Insert reccomended positive / negative prompt
Rating
64 4096
64 4096
1 100
0 30
0 1.5
0 10
1 5
Sampler
Schedule type
Discrete Sampling Type
VAE Model
Prompt Syntax
LoRA1
-8 8
LoRA2
-8 8
LoRA3
-8 8
LoRA4
-8 8
LoRA5
-8 8
Search LoRA for
Sort
Period
Tag
Upscaler
1.1 6
0 512
0 48
0 100
0.1 1
Hires Sampler
Hires Schedule type
-1 30
Adetailer sampler:
Face restoration model
0 1
0 1
Use Textual Invertion in prompt
0.01 1
64 2048
ControlNet model
0 5
0 1
0 1
Preprocessor Name
64 2048
1 255
1 255
0 2
0 20
Model
Mode
0 2
Model
Mode
0 2
Examples
Prompt Negative prompt Images Steps CFG Layer 2 Clip Skip Seed Sampler Img Height Img Width Model

Resources

  • You can also try the image generator in Colab’s free tier, which provides free GPU link.