[Bug/Feature request]: Regional Prompt Latent Mode Compatibility

[metapea]

Checklist

• The issue exists after disabling all extensions
• The issue exists on a clean installation of webui
• The issue is caused by an extension, but I believe it is caused by a bug in the webui
• The issue exists in the current version of the webui
• The issue has not been reported before recently
• The issue has been reported before but has not been fixed yet

What happened?

Forge never got around to fix and add compatibility for Latent in Regional Prompt extensions
like #299 or #242 in sd-webui-regional-prompter

Steps to reproduce the problem

1. Use sd-webui-regional-prompter
2. Click on Latent for generation mode
3. Generate with 2 LoRAs in separated areas from each other

What should have happened?

Generate in there separate area and if there character LoRAs then no details from the character should crossover to a other character while still able to interact like shaking hands.

What browsers do you use to access the UI ?

Mozilla Firefox, Other

Sysinfo

N/A
It's isn't a crash

Console logs

N/A

Additional information

No response

[altoiddealer]

FWIW, there's Forge Couple which should do what those extensions do except better. Although maybe those extensions are supposed to play better with other extensions?

[metapea]

FWIW, there's Forge Couple which should do what those extensions do except better. Although maybe those extensions are supposed to play better with other extensions?

I know about Forge Couple, it's better then sd-webui-regional-prompter with Attention mode. But forge couple doesn't have Latent, just Attention, since Latent still doesn't work on forge.

[metapea]

It is a bit late but "online loras" are currently available now: https://github.com/lllyasviel/stable-diffusion-webui-forge/blob/2f1d04759f36122cbb3553f0f0758755790624c7/backend/operations.py#L14 - lllyasviel

View “online lora” as a module similar to ControlLoRA so that it is moved to GPU together with model when sampling, achieving significant speedup and perfect low VRAM management simultaneously. - lllyasviel