[enhancement]: Generation Presets to Save and Load All Settings (Model, LoRAs, Prompts, etc.) #8640
Description
Is there an existing issue for this?
- I have searched the existing issues
Contact Details
No response
What should this feature add?
Is your feature request related to a problem? Please describe. Yes. Currently, switching between different generation setups (e.g., one style for realistic portraits, another for anime, a third for landscapes) is a manual and time-consuming process.
Each time I switch, I have to manually change:
The checkpoint model
The positive and negative prompts
Multiple LoRAs and their specific weights
Sampler settings, steps, CFG scale, resolution, etc.
Loading metadata from a previously generated image or using the workflow system are possible workarounds, but these methods are often cumbersome and slow (煩わしい - bothersome) when all I want to do is quickly load a frequently used "favorite" setup.
Describe the solution you'd like I propose a new "Preset" or "Template" feature.
This feature would allow a user to save the entire current state of the generation parameters into a named preset (e.g., "My Portrait Style", "Anime Setup v1").
A preset should save and be able to reload all of the following settings with one click:
Positive Prompt
Negative Prompt
Checkpoint Model
All applied LoRAs and their individual weights
All other core settings (Sampler, Steps, CFG Scale, Resolution, etc.)
This would allow users to build a library of their favorite setups and switch between them instantly from a simple dropdown menu, drastically speeding up the workflow.
Alternatives
Loading from Image Metadata: This is useful but requires me to find a specific image that has the settings I want. It's not as fast or organized as selecting a named preset from a list.
Workflows: The workflow system is very powerful, but it can be too complex for this simple use case. A preset system would be a more user-friendly feature for saving and loading common setups quickly.
Additional Content
This feature would save a significant amount of time by eliminating the repetitive manual task of changing models and LoRAs. It would make the tool much more efficient for artists who work with multiple distinct styles.